Prashanth Prakash

1. Snapshot and testable edge thesis

The present verdict is therefore deliberately asymmetric: the record strongly supports institution building, a repeatable vocabulary for looking, and several documented operating relationships; it provides only moderate support for a repeatable Prakash-specific selection edge and insufficient public record for person-level returns. [researcher inference · strong] Portfolio bottom line

2. Core philosophy

Prakash's direct voice begins with preparation rather than prediction. In 2023 he urged investors to pursue new ideas rather than the “flavour of the season,” while describing a prepared mind as the condition for recognizing them. [investor-stated · strong] Mint interview note His 2020 joint interview describes instinct as conditioning by accumulated experience, not a license to ignore evidence. [investor-stated · moderate] Instinct and intuition note

The clearest opportunity sequence is operational. Start with an unmet need; identify a customer the company can reach predictably; bound the initial market; test the wedge; then ask whether adjacent demand can make the company venture-scale. [investor-stated · moderate] Markets & Opportunities note Erasmic's captured 2008 philosophy independently required customer validation, differentiated delivery, comparatively low capital needs, a credible India revenue path, a domain in which the team could help, and working compatibility. [contemporaneous record · strong] archived Funding Philosophy

Treating the founder as a learning system rather than a résumé is a reconstructed rule. [researcher inference · strong] Founder-learning model Captain Fresh's joint investment essay emphasizes domain obsession and the founder's ability to teach the investors about a fragmented supply chain; the BlueStone IPO reflection emphasizes curiosity, local category understanding, endurance, and willingness to change the model. [investor-stated · moderate] Captain Fresh essay · BlueStone reflection Because these are interested, selected accounts, they do not establish comparative predictive power. [researcher inference · strong]

Capital is a means of buying verified progress. The early record favors customer proof, cash conservation, and a plausible sustaining-revenue path; later interviews require a visible first-rupee profit mechanism and financing instruments matched to the business. [investor-stated · strong] 2008 business-plans note · profit mechanism, 19:35–20:38 · venture-debt continuum The reconstructed rule is not “never burn”; it is to specify what evidence each unit of capital must purchase and what result stops the next unit. [researcher inference · strong] Capital-efficiency model

The operating philosophy is collective and bounded. Prakash's current Accel profile explicitly prefers collective credit, the Erasmic pages advertise a four-partner operating service, and multiple cases name other partners as sponsors or governors. [investor-stated · strong] for the norm; [documented behavior · strong] for shared work. Accel profile · Erasmic founding-team archive Mentoring is supposed to be permissioned, but Prakash also reports a 15–18-company active load and limited unilateral board power. [investor-stated · moderate] Master Mentor note

Synthesized as a decision system, the philosophy is: prepare a falsifiable map; enter through a specific reachable customer or coordination bottleneck; select founders who can teach and update; fund evidence rather than narrative; standardize only the binding layer; match capital and governance to the milestone; distribute work across named people; and distinguish paper value from cash. [researcher inference · strong] This synthesis should be used as a checklist to test Prakash's record, not quoted as his own words. Mental-model field guide · deal forensics

3. Sourcing

How opportunities entered the funnel

• Operator and founder networks. Prakash and Subrata Mitra first used friends, industry angels, HNIs, and experience from five early IT deals to form Erasmic. [contemporaneous record · strong] Business Today, paras. 100–105 This is access evidence, not proof that network-sourced deals were better. [researcher inference · strong]
• Prepared thematic search. The record moves from 2008 cross-border services and consumer/mobile opportunity to supply chains, agritech, brands, climate, advanced manufacturing, applications of AI, and longevity. [investor-stated · moderate] Written chronology Theme lists define where to look; they do not establish a company-level thesis or deployed allocation. [researcher inference · strong]
• Founder teaching. Captain Fresh's authors say they tracked Utham Gowda for a year and learned the seafood system from his domain model; QwikCilver's account describes early meetings before stored value was widely understood. [retrospective witness · moderate] Captain Fresh essay · QwikCilver account
• Institutional platform and referrals. The Accel platform supplies a wider partner, founder, executive, policy, and later-stage network than Erasmic could. [researcher inference · moderate] Public records do not disclose source-to-meeting, meeting-to-investment, conflict, or source-to-return denominators. Fund VIII announcement · evidence still missing
• Civic and ecosystem access. Startup-policy, ACT, Bengaluru, and scientific networks can reveal unsolved systems and experts. [researcher inference · moderate] They also require vehicle, information-sharing, related-party, and recusal boundaries; the public record shows no general framework, which is an unanswered governance question rather than an allegation. ACT annual return, pp. 8–11 · civic boundary

Repeatability test

The repeatable portion is not Prakash's personal Rolodex. It is the procedure: date the structural view; name the buyer and bottleneck; identify what the source has directly observed; record conflicts; ask what new fact changes the map; and preserve passes so the funnel can be audited. [researcher inference · strong] Prepared-mind model · reachable-customer model

An IC should reject “proprietary sourcing” when it means only prestige, geography, policy proximity, or a warm introduction. It should ask for the source owner, relationship history, information gained before the financing, other investors who saw the opportunity, the rejected-opportunity set, and any conflict or allocation constraint. [researcher inference · strong] The absence of those denominators is a central limit of the public Prakash record. Source-map coverage

4. Picking

The public record does not disclose a Prakash scoring model, pass rate, reference template, valuation threshold, IC voting rule, veto, conflict rule, or complete post-governance diligence checklist. [researcher inference · insufficient public record] Material-case role matrix The disciplined reconstruction is therefore procedural: freeze the date; state the buyer, bottleneck, and alternative; test founder knowledge; reproduce customer and technical evidence; map unit economics and financing needs; name the responsible partner and dissent owner; set kill criteria; and define the next evidence date. [researcher inference · strong] No-hindsight labs

Red flags are a founder who cannot show learning; inaccessible or subsidized customers; a wedge with no economic adjacency; capital needs that outrun vehicle reserves; IP without qualification; inventory, working-capital, or integration risk treated as “execution”; governance concentration without independent information; and a category thesis standing in for company evidence. [researcher inference · strong] Contradictions and boundaries

5. Deal and portfolio mechanics

What the record supports

• Initial cheque and stage. The Erasmic plan proposed $100,000–$500,000 cheques and intensive work in four to six companies; the 2008 successor planned sub-$1 million seed starts across 25–30 companies. [contemporaneous record · strong] Erasmic Overview · Mint, paras. 133–138
• Follow-on capacity. Larger Accel vehicles made multi-round support possible, and reporting describes more than 40 follow-on companies and a 12–15 year firm-level horizon. [retrospective witness · moderate] Fortune India, paras. 211 and 216 This does not reveal reserve ratios, pro-rata rules, loss budgets, partner allocation, or whether Prakash made each decision. [researcher inference · strong]
• Governance. BlueStone filings document Prakash's director role. [contemporaneous record · strong] BlueStone prospectus, pp. 24–25 and 147–148 Infra.Market identifies an India V nominee channel. [contemporaneous record · strong] Infra.Market role record RentoMojo identifies him as a non-executive Accel nominee director. [contemporaneous record · strong] RentoMojo DRHP, p. 235 Board presence proves neither sourcing nor intervention efficacy. [researcher inference · strong]
• Liquidity. The public record supports a partial BlueStone offer-for-sale, an exact UrbanTouch vehicle cost and sale record, and headline transaction consideration for several acquisitions. [contemporaneous record · strong] Economic ranking It rarely supplies seller allocation, remaining value, fees, taxes, timing, or distributed cash. [researcher inference · strong]

What must remain unknown

No universal ownership target, entry-price rule, security preference, anti-dilution policy, board-right threshold, reserve policy, follow-on scorecard, concentration limit, sell discipline, or partner-level carry record is public. [researcher inference · insufficient public record] Vehicle and economics retrieval ledger The 52 researched opportunity groups are a searched candidate set, not a career denominator: the current 28-company graph is mutable, Erasmic's 15 investments are only partially named, and the Padma wording is not an enumerated roster. [researcher inference · strong] Reconciliation

Portfolio construction should therefore be inferred only at the level the evidence permits. Early Erasmic concentration plausibly matched promised operating intensity; larger Accel funds widened breadth and follow-on capacity but also created deployment, attention, escalation, and exit-pressure risks. [researcher inference · moderate] Dimension-matched fund comparison BlueStone dominates observable Prakash-linked cash evidence, while UrbanTouch is the strongest exact negative cash control and Good Glamm the strongest operating/governance failure control. [researcher inference · strong] Pareto analysis

For modern IC use, treat every follow-on as a new decision: compare milestone delta with the original memo, reprice the remaining market and financing risk, record outside validation, calculate reserve opportunity cost, assign an independent kill advocate, and vote separately on company quality and vehicle fit. [researcher inference · strong] Patient-capital model A board seat should have named information rights, intervention scope, time budget, conflict controls, and a handoff or exit condition. [researcher inference · strong] Mentoring-capacity model

6. Vehicle and incentive fit

The incentive conclusion is bounded. Small early vehicles could align concentration with labor; larger institutional funds could finance long learning arcs and follow-ons; parallel entities could allocate stage and liquidity differently. [researcher inference · moderate] Yet no public LP agreement shows duration, fundraising pressure, management fees, carry, recycling, reserves, stakeholder votes, removal, succession, or key-person continuity. [researcher inference · insufficient public record] Full vehicle ledger

The strategy therefore cannot be evaluated independently of vehicle fit. A locally specific, slow-qualification company may be attractive for a vehicle with sufficient duration and reserves but destructive for one facing deployment or liquidity pressure; an intensive mentoring promise may work in a four-to-six-company pool and become marketing language across a much larger partner workload. [researcher inference · strong] Fund-scale model · capacity model

7. Team attribution and work with founders

The public record is incompatible with a lone-picker biography. Erasmic presented four partners—Subrata Mitra, Prakash, Mahendran Balachandran, and Gagan Kumar—with complementary operator, engineering, sales, and management experience. [contemporaneous record · strong] Erasmic founding-team archive Later Accel cases name still more partners, founders, executives, vehicle directors, and nominees. [documented behavior · strong] People map

Prakash's stated mentoring norm is to help by permission rather than force. [investor-stated · moderate] Mobile-first note That norm does not remove financing or information asymmetry: board rights, preferred securities, runway, follow-on control, and distressed-company leverage can constrain founder choice. [researcher inference · strong] Founder-choice contradiction

The operating promise also has a capacity limit. Erasmic advertised help across strategy, customers, organization, technology, operations, marketing, and finance; Prakash later described 15–18 active companies. [contemporaneous record · strong] for the promised service; [investor-stated · moderate] for the workload. Erasmic value add · Master Mentor note A credible support plan therefore needs a named owner, founder consent, hours and specialist cost, baseline, intervention, milestone, handoff, and portfolio opportunity cost. [researcher inference · strong] Permissioned-mentoring model

8. Views on the craft

• Power law and outliers. Prakash does not publish a formal power-law doctrine. The observable record is nevertheless outlier-sensitive: BlueStone dominates attributable cash evidence, while the complete portfolio denominator and personal cash flows are unavailable. [researcher inference · strong] Economic ranking A great investor should preserve upside without letting one winner launder the ordinary-case record. [researcher inference · strong]
• Fund size. Erasmic's small pool paired low cheque sizes with intense operating support; later Accel vehicles paired seed entry with follow-on breadth. [contemporaneous record · strong] Vehicle comparison Bigger is useful only if opportunity count, ownership, reserves, partner bandwidth, and exit markets preserve return math. [researcher inference · strong]
• Access versus selection. The record documents founder and operator networks. [contemporaneous record · strong] Sourcing evidence It also documents Accel-platform reach, Bengaluru ecosystem participation, and civic roles. [documented behavior · strong] Fund VIII platform · Bengaluru access audit · Civic-role boundary Taken together, these channels are potential access advantages. [researcher inference · strong] They are not selection skill until a complete funnel shows which accessed deals were chosen, passed, conflicted, or lost and what happened. [researcher inference · strong]
• Governance. Prakash celebrates mentoring and reports limited unilateral board control; public filings prove some nominee roles but not interventions or their causal outcomes. [investor-stated · moderate] 2023 new-ideas note · governance evidence Great governance is an information and decision-rights system, not attendance or proximity. [researcher inference · strong]
• Exits. The 2018 record separates paper marks, stock consideration, and distributable cash; later talks treat domestic IPOs and M&A as ecosystem infrastructure. [retrospective witness · strong] Fortune India, exit discussion · liquidity framework A financing round, valuation, unicorn label, acquisition headline, or IPO filing is not a realized return. [researcher inference · strong]
• Great-investor standard. The most defensible standard in this record is to develop a prepared but revisable view, identify a locally real bottleneck, learn from founders without surrendering verification, match evidence to capital, distribute credit accurately, preserve governance capacity, and realize cash without forcing the wrong clock. [researcher inference · strong] Mental models The public record does not prove Prakash met that standard across a denominator; it shows why the standard is useful. [researcher inference · strong] Bottom line

9. Signature bets and decision process

The six decision labs preserve the information freeze, ask the analyst to decide, and reveal the actual decision and outcome only afterward. They should be used as exercises, not retold as hindsight-supported “proof” of a philosophy. [researcher inference · strong] No-hindsight protocol

The replay discipline is the process lesson. Before each reveal, write: invest or pass; the alternative thesis; the exact kill criteria; acceptable security, ownership, and financing exposure; the responsible partner and dissent owner; and the next-check milestone. [researcher inference · strong] Afterwards, score the original reasoning separately from the outcome so luck, price, governance, and subsequent decisions do not collapse into one verdict. [researcher inference · strong] Decision-learning ranking

10. Mental models and maxims

The field guide contains 15 distinct supported models. The labels below describe the origin of each rule; they do not assert that the rule generated excess return. [researcher inference · strong] Evidence-gap inventory

The maxims compress to three safeguards: a prepared mind must still be disconfirmed; patience must still be gated; and hands-on help must still be attributed, permissioned, and capacity-priced. [researcher inference · strong] Each model's full entry supplies mechanism, procedure, examples, counterexample, evidence breadth, and a dated transfer test. Mental models

11. Evolution and contradictions

Four durable tensions should remain visible. Intuition tells the investor where to look, while evidence decides whether to proceed; Prakash never states that reconciled formula in one source. [researcher inference · moderate] Prepared-mind model Patient capital can preserve learning, while finite funds and reserves can reward escalation; no reserve ledger resolves the tension. [researcher inference · strong] Patient-capital model Permissioned mentoring respects founders, while preferred capital and board rights still create power. [researcher inference · strong] Capacity model Finally, ecosystem participation can improve problem access while increasing disclosure and recusal needs; no general conflict framework was found. [researcher inference · strong] Civic boundary

12. Failures and limits

The largest limitation is the denominator. The 52 researched opportunity groups are neither all opportunities seen nor all investments made; no complete passes, write-offs, ownership histories, reserve decisions, partner allocations, or fund cash flows are public. [researcher inference · strong] Evidence boundary Public narratives disproportionately explain winners after milestones and recent themes before outcomes; this creates survivorship, hindsight, and advocacy bias. [researcher inference · strong] Writings evidence boundary

Era dependence is also material. A 2004–08 seed-capital gap, Bengaluru's operator network, India's mobile and payments build-out, global liquidity, and later domestic IPO capacity cannot be assumed in another geography or vintage. [researcher inference · strong] Historical context The transferable object is a diagnostic method, not the historical sector list, cheque size, holding period, or founder network. [researcher inference · strong]

13. Historical operating environment

Available base rates are sparse. The public sources permit some market and regulatory context but no matched India seed cohort with consistent vintage, stage, ownership, write-off, and distributed-cash data. [researcher inference · insufficient public record] Denominator boundary Later IPOs, acquisitions, and current profile cards are outcome-selected and cannot be imported into an original decision packet. [researcher inference · strong] The labs therefore use the best dated company and market evidence available, then expose the reveal separately. Decision labs

14. Modern VC translation (as of 2026-08-20)

Every row below is an analyst-created transfer test. Historical evidence establishes the source rule; the current procedure and misuse warning are [researcher inference · strong] unless a row says otherwise.

The translation is intentionally stricter than the historical narratives. The public record shows a $650 million India VIII vehicle. [contemporaneous record · strong] Fund VIII Form D Prakash's 2025 discussion spans equity, venture debt, working capital, and asset finance. [investor-stated · moderate] Venture-debt note Current AI discussion includes cheaper voice interfaces and application-layer opportunities. [investor-stated · moderate] Voice-first AI note Current manufacturing work describes longer qualification and gestation even when more capital is available. [investor-stated · moderate] Manufacturing gestation Those conditions increase the value of explicit milestones, attribution, capacity accounting, and vehicle-fit votes. [researcher inference · strong] They do not establish that Prakash or Accel use this complete analyst-created checklist. [researcher inference · strong]

15. Comparison with completed peers

Peter Brooke is a valid institution-builder and operator-to-investor comparator: he repeatedly changed vehicles, emphasized local information and operating support, and confronted the capacity and authority problems of cross-border investing. Arthur Rock is a valid people-first and governance comparator: he used relationship access, behavioral founder judgment, flexible structures, long board involvement, and explicit attention to negative information. [researcher inference · strong] Both packages are complete; no unfinished peer is used. Different eras, geographies, vehicle forms, currencies, and return methods make this a craft comparison, not a performance ranking. [researcher inference · strong]

The sharpest disagreement is about where the reusable edge sits. Brooke's record locates it in institution and vehicle design across local contexts. [researcher inference · moderate] Brooke edge Rock's locates it in people judgment plus governance information. [researcher inference · moderate] Rock edge Prakash's locates it in local bottleneck recognition plus patient platform support. [researcher inference · moderate] Prakash edge No common denominator establishes which mechanism contributes most to return. [researcher inference · strong] Bounded comparison

16. The strongest case against greatness

Skeptical thesis. [researcher inference · strong] Prakash may be a consequential institution builder and ecosystem participant whose investing reputation reflects fortunate presence in Bengaluru as Indian venture professionalized, privileged access to founders and policy networks, Accel's brand and follow-on capital, shared partner work, selected winner narratives, and one dominant observable cash case—without a full denominator proving superior repeatable personal investment performance.

1. Access may explain the funnel. Prakash's account describes family resources and early workshop and computer exposure. [investor-stated · moderate] FCTC part 1, 12:29–33:43 The record also documents operator, Bengaluru, Accel-platform, and civic access. [documented behavior · strong] Access-versus-skill audit Whether those channels explain access independently of selection is a skeptical hypothesis, not his stated conclusion. [researcher inference · strong]
2. Team attribution is central. Erasmic had four partners; QwikCilver, MyGlamm, and Captain Fresh are joint accounts; transaction-specific records assign Flipkart, Swiggy, Freshworks, Ola, and other work elsewhere. [documented behavior · strong] People map
3. The denominator is missing. The 52 researched groups are a constructed public candidate set, not all looks, investments, write-offs, or cash flows; no pass ledger, ownership series, reserve schedule, or person-level return record exists. [researcher inference · strong] Reconciliation
4. Observable economics are concentrated. BlueStone is the strongest attributable cash case, while UrbanTouch supplies an exact negative vehicle control and many famous outcomes cannot be assigned to Prakash. [researcher inference · strong] Economic ranking
5. Public stories select winners and recent theses. IPO essays and investment announcements explain successes retrospectively; manufacturing, AI, and longevity writing advocates themes before mature outcomes. [researcher inference · strong] Writings coverage
6. Ordinary and failed cases challenge the screen. Teabox remained ordinary; BabyOye and UrbanTouch lost value; Good Glamm collapsed despite celebrated resilience and distribution; climate tech underdelivered. [researcher inference · strong] Failures
7. Era and vehicle effects are powerful. Scarce seed capital, India's mobile and payment build-out, abundant global capital, Accel reserves, and domestic liquidity could reward presence and patience without proving company-selection skill. [researcher inference · strong] Historical environment
8. The stated system is partly reconstructed. Prepared mind, customer wedge, founder learning, capital efficiency, and platform support recur, but Prakash never publishes the complete combined checklist, scoring weights, kill thresholds, or reserve rules. [researcher inference · strong] Mental models
9. Process changes remain unvalidated. Nearly doubled diligence time and sharper manufacturing thresholds may be sensible reactions, but elapsed time and a new framework are not evidence of better detection or returns. [researcher inference · strong] Contradictions

The strongest favorable reply is narrower than a personal-alpha claim. Prakash helped create an early seed institution when small cheques and operating support were scarce; the record shows several locally specific category relationships, a documented long BlueStone board role, willingness to preserve some negative admissions, and a later vocabulary that is more instrumented than the earlier thematic optimism. [researcher inference · moderate] Profile edge test That makes the record a useful decision curriculum even before it becomes a verified performance table. [researcher inference · strong]

Evidence would favor the favorable thesis if dated IC memos, passes, partner allocations, reference work, cap tables, intervention logs, votes, follow-ons, and owner-specific cash flows show that the proposed mechanisms were present before outcomes and improved ordinary-case results after controlling for access, team, vehicle, vintage, and outliers. [researcher inference · strong] Evidence would favor the skeptic if roles are mostly institutional or nominal, category enthusiasm substitutes for company evidence, follow-ons reflect escalation, ordinary-case cash returns are weak, or results concentrate in founders and market regimes available to Accel regardless of Prakash. [researcher inference · strong] Primary records that could change the assessment

17. Analyst study guide, glossary, and go deeper

Recommended reading path

1. Start with the biography's falsifiable edge. Write the edge, mechanism, scope, alternative explanation, prediction, and falsifier in your own words.
2. Read the 2008 Erasmic doctrine, then the written chronology. Mark which claims are firm voice, joint writing, sole byline, or researcher synthesis.
3. Read the spoken frameworks and contradictions. Separate current aspiration from demonstrated process change.
4. Complete all six no-hindsight labs before each actual-decision reveal. Always include the negative UrbanTouch lab and the Good Glamm failure lab.
5. Reconcile the 52-group ledger, role matrix, and vehicle ledger. Reject any person-level conclusion that cannot survive all three.
6. Use the 15-model field guide to convert the record into IC procedures, then reread the skeptical case.
7. Compare only the completed peers: Peter Brooke for institution/vehicle design and Arthur Rock for people/governance architecture.

Investor-specific glossary

Unless a row carries a stronger source label, each definition is [researcher inference · moderate] and applies only within this package.

Reusable sourcing checklist

• What structural view was dated before the company appeared, and what would disconfirm it?
• Who introduced the opportunity, what did that person directly observe, and what incentive or conflict did they have?
• Which customer is reachable now, through what channel, at what cost, and with what behaviorally costly proof?
• What local or system-specific bottleneck exists beyond a fashionable sector label?
• Which comparable opportunities were seen, passed, conflicted, or lost, and are those outcomes preserved?
• Does civic, policy, university, founder, or portfolio proximity require disclosure, information separation, or recusal?

Reusable diligence and IC checklist

• Freeze the date; separate known, unknown, disputed, inferred, and later-reveal facts.
• Ask the founder for one reproducible operating insight, one changed decision, one failed experiment, and one expert gap.
• Independently test customer behavior, contribution, collections, technical performance, qualification, data rights, and regulation.
• State the base case, strongest alternative, kill criteria, and next evidence date.
• Specify security, ownership, dilution, check, reserves, follow-on threshold, and fund-return fit; mark unavailable inputs explicitly.
• Name sourcing, sponsorship, diligence, approval, negotiation, governance, recruiting, follow-on, dissent, and exit owners; keep unproved roles unknown.
• Price the operating-support plan in partner hours and specialist cost; record founder consent and handoff.
• Vote separately on company quality, price, vehicle fit, conflict, and governance capacity.

Reusable governance and follow-on checklist

• What information rights reach the board independently of the founder narrative?
• Which metrics, covenants, technical gates, or customer signals require intervention, resizing, or stop?
• Who owns dissent, and can that person access customers, finance, product, legal, and people evidence?
• What changed since the original memo, and is it evidence or merely more capital and elapsed time?
• What is the reserve opportunity cost relative to new investments and stronger portfolio companies?
• Are founder choice, financing leverage, board duty, conflicts, and recusal documented?
• Does the vehicle have enough duration and liquidity tolerance for the next milestone?
• At exit, separate offer event, sold stake, gross proceeds, fees/taxes, remaining value, and LP distribution.

Go deeper — ranked sources

The ranking is [researcher inference · moderate] and prioritizes decision utility, directness, adverse value, and coverage rather than praise.

1. Erasmic's captured 2008 pages — the earliest surviving institutional screen, strategy, vehicle proposal, operating-service claim, team record, and track-record boundary; unbylined team voice, not Prakash-only doctrine.
2. Mint — “Invest in new ideas, not flavour of the season” — the clearest prepared-mind, anti-fashion, board-power, valuation, exit, and LP-liquidity discussion.
3. SeedToScale — Markets & Opportunities — the most reusable customer-access, unmet-need, bounded-TAM, adjacency, and heavy-lifting sequence.
4. BlueStone prospectus — the strongest primary control for vehicle, security-holder, board, dilution, and partial-liquidity analysis.
5. Mint — previous misses and manufacturing reset — the strongest adverse thematic update and the right control against treating the new manufacturing framework as already validated.
6. Fortune India — The Accel Effect — useful for follow-on breadth, reported duration, weak/consolidated bets, and the distinction between paper exits and cash; still a firm narrative rather than audited fund data.
7. Master Mentor — the best career, luck, mentoring, workload, and public/private-boundary narrative, with retrospective limitations kept visible.
8. Tribunal UrbanTouch record — the strongest exact negative vehicle cash control and a model for refusing transaction rumors.
9. Advanced Manufacturing report — the clearest current IP/readiness/qualification screen; a joint, interested, pre-outcome framework.
10. Source map — the 220-record denominator, archive routes, reliability priorities, access states, and unresolved primary records needed to audit every synthesis claim.

The final exercise is to write two one-page IC memoranda from the same frozen case: one using the favorable edge thesis and one using the skeptical thesis. Both must name the buyer, bottleneck, founder-learning proof, capital gates, vehicle fit, attribution map, kill criteria, and evidence that would change the decision. [researcher inference · strong] If the two memos differ only in tone, the analyst has not operationalized the record. Decision labs

Pareto 80/20 — what drove the record

The public record does not support a literal 80/20 calculation. There is no complete Prakash-personal portfolio, security-level cash-flow ledger, allocation record, carry schedule, or audited fund-return series. The rankings below therefore separate cash-evidence proxies, historical importance, and decision-learning value; they do not equate company fame, acquisition price, IPO proceeds, or board service with Prakash's personal return. [researcher inference · strong] Evidence boundary

Economic contribution — ranked by attributable cash evidence, not fame

Removing BlueStone leaves UrbanTouch as the only associated case with public vehicle cost and sale consideration, although its Prakash link remains provisional; the other transactions supply noncomparable headline consideration. A concentration test excluding BlueStone is therefore not computable. [researcher inference · strong] Portfolio-level economic boundary

Historical importance — ranked independently of economics

Decision-learning value — ranked for replayability

Evidence boundary and searched denominator

Scope, inclusion, and deduplication

• A company family enters the named Prakash-linked set only when an official Accel relationship page, contemporaneous Erasmic record, or direct/joint investment account associates Prakash or his then-current team with capital, a formal portfolio, or a clearly investment-framed relationship. One company family is counted once across renames, rounds, follow-ons, mergers, and exits. [researcher inference · strong] Official profile graph · Erasmic merger record
• The sweep covers the mid-1990s through 2026 and the variants Prashanth Prakash, Prashant Prakash, Prakash Prashanth, Erasmic/EVF/Accel India, company aliases, and vehicle names. Routes included the live Accel graph, SEC/SEBI/IBBI and court/tax filings, government FDI and award records, company releases, fund announcements, press archives, DIN/MCA-derived leads, Common Crawl recovery, Wayback attempts, and transaction/acquirer records. [researcher inference · strong] source-map identity/search scope · archive/access ledger
• Current website association proves only Accel's present label. It does not by itself prove sourcing, sponsorship, diligence, approval, check ownership, board work, exit influence, or personal economics. [researcher inference · strong] Accel profile and relationship cards
• Prakash personally, the 2004 investment club, Erasmic Venture Fund, each Accel India LP, Mauritius holding companies, Accel growth vehicles, co-investors, portfolio-company boards, and ACT's philanthropic capital are separate actors. [researcher inference · strong] Vehicle sequence
• A company acquisition price, IPO market capitalization, disclosed selling-shareholder proceeds, or private valuation is not an investor return. A return multiple or IRR is not calculated without the relevant security, ownership, dilution, cash-flow timing, fees, and distributions. [researcher inference · strong] BlueStone prospectus, cover and capital structure

Reconciliation

The reproducible named set contains 52 opportunity groups: 28 current official Accel relationship pages, eight additional Erasmic/EVF names, two Padma-named noncurrent outcomes, two personal/non-Accel-looking investments, two shared essay ties, seven additional board/transaction cases, and three team/profile cases retained to test attribution. It is a discovery denominator, not a funded-career, ownership, or return denominator. [researcher inference · strong] Current graph · Additional ledger

Coverage fraction: not computable. The numerator can be named as 52 researched candidates, but the career denominator cannot: the Padma citation first names eight bets and then says “over 50 other companies,” without an enumerated roster or consistent scope; the 15 Erasmic investments are only partially named; the current 28-card graph is mutable and survivorship-biased; and private/personal positions are not disclosed. Neither 52/>50 nor 52/>58 is a defensible fraction. [researcher inference · strong] Padma citation, p. 1 · Current graph · 2008 Erasmic count

Historical market, regulatory, and base-rate context

• In 2004, the investable context described by the founders was not a mature institutional seed market: Prakash and Mitra first pooled friends', angels', and HNIs' capital in an investment club, then converted experience from five early deals into a formal $10 million fund. [contemporaneous record · strong] Business Today, paras. 100–105
• At the July–November 2008 combination and close, the team described Indian early-stage capital as underserved even though later-stage firms were present; the $60 million successor closed after the global market break, used mostly institutional LPs, planned sub-$1 million initial seed checks, and targeted 25–30 companies. [contemporaneous record · strong] Business Today, paras. 105–107 · Mint, paras. 133–138
• Period SEBI rules defined a domestic venture fund as a dedicated private pool, required disclosure of strategy, capped a single undertaking at 25% of corpus, and generally required 75% of investible funds in unlisted equity or equity-linked instruments. [contemporaneous record · strong] SEBI Venture Capital Funds Regulations, regs. 2, 11–12 These rules are context, not proof that every offshore Erasmic or Accel entity used that domestic form. [researcher inference · strong] Accel India IV Form D, items 1 and 4
• A period academic review describes 2004–08 Indian private-equity/venture activity as heavily foreign-funded, biased toward later stages, and shortened by the boom-and-contraction cycle. [retrospective witness · moderate] India Venture Capital and Private Equity Report 2009 That is a market comparator, not a base rate for Prakash's selection skill or returns. [researcher inference · strong] Evidence boundary
• The 2010s cases reflect expanding consumer internet, payments, logistics, and marketplace infrastructure, but each encountered category-specific physical constraints: QwikCilver needed retailer trust and RBI authorization; BlueStone needed authenticity, touch, exchange, inventory, and manufacturing controls; Infra.Market needed procurement, quality, working capital, and distribution. [researcher inference · strong] QwikCilver account · BlueStone deck · Infra.Market discussion
• The 2021 MyGlamm thesis was published during a capital-rich roll-up period; the later Good Glamm record includes a flat round, defaults, arbitration, delayed salaries, a board resignation, and lender-driven asset sales. [retrospective witness · strong] Economic Times timeline Later distress cannot be imported into the 2021 decision freeze, but it is necessary evidence against treating growth, content distribution, or resilience as sufficient. [researcher inference · strong] MyGlamm essay

Vehicles, economics, and incentive boundaries

Dimension-matched modern-fund comparison — as of 2026-08-20

People, capital, and influence map

Material-case role matrix

Unknown below means the inspected public record does not assign that function; it does not mean nobody performed it. Website partner labels and board seats do not silently fill sourcing, IC, or economic fields. [researcher inference · strong] official relationship graph

Comprehensive public deal and opportunity ledger

Current official relationship graph — 28 company families

Each row below reports Accel's current website fields as of 2026-08-20. “No exit label” means only that the inspected Accel page did not display one; it does not prove the company remains active, independent, valuable, or held. [researcher inference · strong] Prakash profile relationship graph

Additional historical, personal, board, and team cases — 24

Exclusions and attribution controls

• NetKraft and VisualReality are operating-company chapters, not investments made by Prakash. NetKraft's all-cash sale had undisclosed value after $8.5 million raised; it is formative operator evidence, not a venture return. [contemporaneous record · strong] Rediff/PTI, paras. 42–53
• ACT's environment/health grants, IISc Longevity India support, and other philanthropy are separate from Accel capital. [contemporaneous record · strong] ACT About · IISc launch
• S4S is explicitly discussed as a company in which the speaker says Accel had not invested; an unnamed materials-company intention is not a closed deal. [investor-stated · moderate] TIA Summit note, 21:43–27:21
• Zetwerk and Fashinza appear as Accel investments in Prakash's manufacturing essay, but the essay does not say he sourced, sponsored, approved, governed, or owned them. [investor-stated · moderate] Moneycontrol, B2B marketplace section
• HolidayMe is not a second opportunity group: Traveazy describes HolidayMe's 2014 origin, Accel's 2016 Series A, and the later umrahme launch within the same group lineage. [contemporaneous record · strong] Traveazy timeline · Accel umrahme page

No-hindsight decision labs

The exercise prompts in each lab are analyst-created and deliberately appear before every sourced capital response, governance outcome, sale, shutdown, or later lesson. Missing terms, dissent, ownership, votes, and alternatives remain missing; they are not reconstructed from the result. [researcher inference · strong] Evidence boundary

Decision lab 1: QwikCilver — creating a stored-value category

Decision freeze A — 2006 founder meeting. A later Accel account recalls Mahendran Balachandran and Prakash meeting the founders when Indian gift cards were a tiny, poorly understood market and retailers needed education, trust, and infrastructure. Because this description was published years later, it is participant memory of the early information set rather than a contemporaneous memo. [retrospective witness · moderate] Accel India, opening and “Finding founder-market fit”

Decision freeze B — 2010 capital window. No decision-date memo, market sizing, diligence, terms, ownership, approval roles, dissent, reserve plan, or milestones survived in the public record. A reader must therefore decide using only a proposed enterprise stored-value rail, retailer integration/education requirements, and a regulatory authorization path. [researcher inference · strong] Accel company page · retrospective category account

• A prudent contemporary could fund a narrowly scoped retailer-integration proof, wait for authorization and repeat-use evidence, or pass because education cycles could outrun the category. These are analyst-created alternatives, not documented IC debate. [researcher inference · moderate] Accel India account

Reader exercise — decide before revealing the actual decisions or outcome. [researcher inference · strong]

1. Would you invest then? Choose invest, stage, or pass using only retailer willingness, authorization path, integration cost, and founder evidence available at each freeze.
2. What would kill the deal? Specify maximum implementation time, authorization delay, customer concentration, fraud loss, and cash-burn thresholds.
3. What ownership/terms fit the risk? Propose a security, governance rights, ownership target, and reserve cap without assuming a later sale.
4. What evidence unlocks the next check? Require activated retailers, stored-value volume, repeat use, gross margin, fraud loss, integration time, and regulatory milestones.

Actual decisions and outcome. Accel's page records a 2010 initial investment. The later account says the company raised $20 million across Accel and others and jointly names Balachandran and Prakash in the original meeting, but supplies no allocation, board minutes, or tranche logic. Pine Labs confirms the acquisition, and Mint/PTI reported a $110 million price. No source discloses Accel proceeds or Prakash economics. [retrospective witness · moderate] Accel page · Accel account · Pine Labs · Mint/PTI

Decision lab 2: BlueStone — underwriting trust in jewelry commerce

Decision freeze — 2011 deck and Series A decision. The surviving company deck argued that high ticket values and margins, movement toward branded/lightweight fashion jewelry, and a nascent online channel created an opportunity; it proposed a low-inventory start and later category expansion. These were company claims, not independent diligence or an Accel memo. [contemporaneous record · strong] 2011 deck, PDF pp. 5–23

• Jewelry required authenticity, sizing, touch, exchange, logistics, and working-capital trust; the deck's channel and margin claims could be tested before a large rollout. [contemporaneous record · strong] deck, market and model slides
• The exact Accel check, instrument, entry price, ownership, IC sponsor, dissent, alternatives, and intended board plan are unavailable. [researcher inference · insufficient public record] Accel company page
• An analyst could require conversion after certification disclosure, return/exchange behavior, inventory turns, repeat/referral demand, customer-acquisition payback, and evidence that full-stack manufacturing improves quality enough to justify capital intensity. This is a reconstructed test, not a documented term sheet. [researcher inference · moderate] deck

Reader exercise — decide before revealing the actual decisions or outcome. [researcher inference · strong]

1. Would you invest then? Decide whether online trust and margin potential offset inventory, authentication, and return risk.
2. What would kill the deal? Set limits for inventory turns, gross-margin leakage, returns, fraud/authenticity failures, CAC payback, and financing needs.
3. What ownership/terms fit the risk? Specify staged funding, information rights, inventory covenants, governance, and reserve ceiling.
4. What evidence unlocks the next check? Require cohort repeat, branded demand, delivery success, exchange satisfaction, contribution margin, and working-capital discipline.

Actual decisions and outcome. Accel's page records a 2011 Series A. The annual report dates Prakash's board appointment to 26 October 2011; later filings show India III/VII exposure, though not every follow-on decision or intervention. The 2025 prospectus records India III selling 2,603,915 shares at ₹517 for ₹1,346.22 million gross while retaining 13,540,055 shares. The filing does not support a full MOIC, net fund return, or personal return. [contemporaneous record · strong] Accel page · annual report, PDF pp. 17–19 and 125–128 · prospectus, cover and PDF pp. 24–25

Decision lab 3: BabyOye — underwriting specialist commerce and operating support

Decision freeze — 2011 financing. A contemporary report describes a proposed $2.5 million round for a baby-products e-commerce company and a category thesis built around fragmented specialist supply and online convenience. [contemporaneous record · moderate] Medianama The original memo, unit-economics cohort, check allocation, security, ownership, IC vote, dissent, reserve rule, and decision-date assessment of warehousing, hiring, and operating design are absent. [researcher inference · strong] evidence boundary

• Inventory turns, fulfillment reliability, repeat behavior, customer-acquisition payback, supplier terms, and contribution margin are analyst-created diligence tests prompted by the disclosed warehouse/operations model; the inspected sources do not say the IC used them. [researcher inference · strong] founder operating account

Reader exercise — decide before revealing the actual decisions or outcome. [researcher inference · strong]

1. Would you invest then? Decide whether specialist selection and trust can overcome inventory, fulfillment, and acquisition costs.
2. What would kill the deal? Set limits for repeat rate, contribution margin, dead inventory, fulfillment failure, CAC payback, and burn per city.
3. What ownership/terms fit the risk? Choose milestone tranches, inventory controls, board rights, liquidation protection, and a follow-on ceiling.
4. What evidence unlocks the next check? Require cohort retention, private-label margin, inventory turns, supplier terms, on-time fulfillment, and cash conversion.

Actual decisions and outcome. Accel participated in the 2011 round and the founder credited Prakash with operating help. A 2013 report described $12 million of fresh capital and estimated that Accel, Tiger, and Helion together held about 60%, roughly equally; that is press reporting, not a cap table. Mahindra officially acquired the business in 2015, and later reporting characterized the transaction as a likely major VC write-off. Exact consideration, Accel recovery, Prakash economics, board response, stop rule, claimed lesson, and later process change are not public; earlier staging might or might not have improved recovery. [retrospective witness · moderate] 2011 founder account · Times of India · Mahindra release · Economic Times outcome

Decision lab 4: UrbanTouch — setting financing limits in early e-commerce

Decision freeze — 2011 capital/governance window. No accessible contemporaneous investment thesis or memo was located. The only defensible ex-ante framing is that UrbanTouch/Reindeer operated an e-commerce property during India's early online-retail buildout; any more specific rationale would be hindsight invention. Public records do not disclose the original forecast, unit economics, alternatives, approval, dissent, reserve rule, or decision-date downside valuation. [researcher inference · insufficient public record] tribunal record

• Because the information set is thin, the exercise begins by defining a minimum diligence packet rather than presuming a positive thesis. [researcher inference · strong] Evidence boundary

Reader exercise — decide before revealing the actual decisions or outcome. [researcher inference · strong]

1. Would you invest then? List the minimum evidence required before making any decision; do not infer quality from a future buyer.
2. What would kill the deal? Define gross-margin, repeat, inventory, acquisition-cost, cash-conversion, and runway thresholds.
3. What ownership/terms fit the risk? Specify staged financing, preference, information/board rights, and an absolute reserve limit.
4. What evidence unlocks the next check? Require cohort contribution margin, inventory liquidation value, supplier terms, repeat purchase, and runway without a new round.

Actual decisions and outcome. A reproduced tribunal record identifies Accel India Venture II (Mauritius) Limited and records ₹58,752,280 invested across 1,000 equity shares, 36,500 Series A shares, and 8,333 Series B shares as of 1 April 2012. It reports ₹35,017,255 of Accel sale consideration within ₹122.335 million total consideration, describes losses and shutdown, and rejects a reported $30 million sale rumor. Sale consideration divided by the cited investment is approximately 0.596x, a ₹23,735,025 shortfall before fees, taxes, timing, and any unobserved distributions; this is a single-source observed sale-consideration / cited-investment calculation from the reproduced court figures, not a complete fund MOIC, return, or Prakash economics. Board-level votes, interventions, bridge decisions, claimed lesson, and later process change remain unknown, so whether an earlier stop improved recovery is unknowable. [contemporaneous record · moderate] tribunal record, investment and sale findings

Decision lab 5: MyGlamm — underwriting a content-to-commerce roll-up

Decision freeze — 2021 operating thesis. In a later jointly authored account, Anand Daniel, Prakash, and Rachit Parekh argued that MyGlamm's pivot, owned distribution, content-to-commerce engine, and house-of-brands strategy could build a digital consumer platform. The source is a later firm essay, not a decision-date IC memo, and supports joint thesis attribution only. [contemporaneous record · strong] Accel essay, byline and thesis

• The company had survived a prior model change and was using content and acquisitions to aggregate demand and brands. [contemporaneous record · strong] Accel essay
• Acquisition-level unit economics, deferred consideration, integration capacity, debt limits, working capital, salary/vendor obligations, downside runway, vehicle, ownership, decision rights, and dissent are absent from the essay. These are uncertainty fields, not evidence that the risks had already materialized. [researcher inference · strong] Accel essay

Reader exercise — decide before revealing the actual decisions or outcome. [researcher inference · strong]

1. Would you invest then? Decide whether distribution and content synergies justify a roll-up without assuming successful refinancing.
2. What would kill the deal? Set limits for brand-level contribution loss, integration slippage, deferred liabilities, debt covenant stress, salary/vendor delays, and runway.
3. What ownership/terms fit the risk? Propose acquisition approval rights, debt limits, reporting covenants, milestone tranches, and liquidation protection.
4. What evidence unlocks the next check? Require organic versus acquired growth, brand-level cash generation, cross-sell, working capital, covenant headroom, and a no-refinancing survival plan.

Actual decisions and outcome. Accel invested under the joint 2021 thesis and reportedly joined a $30 million flat round in 2024; available reporting identifies Anand Daniel, not Prakash, as Accel's director who resigned in December 2024. [retrospective witness · moderate] original thesis · Economic Times timeline · board report The NCLT admitted Sanghvi Beauty & Technologies to CIRP on 12 September 2025 after documented debt/default and inability to pay. [contemporaneous record · strong] NCLT order, paras. 6–9, 15–28 and 38–45 No inspected evidence assigns Prakash the acquisitions, debt, bridge, board resignation, or insolvency decisions, and no Prakash-specific postmortem or process change was found; stronger acquisition gates might have helped, but the counterfactual is unobserved. [researcher inference · strong] failure record

Decision lab 6: Deep Rooted — underwriting a perishables channel pivot

Decision freeze A — 2018. No contemporaneous memo, check, instrument, ownership, approval, dissent, or milestone plan was recovered. A reader should treat the original greenhouse/perishables rationale as unavailable at this first freeze rather than import a later description. [researcher inference · insufficient public record] evidence boundary

Decision freeze B — 2020 capital window. A period company account presented a greenhouse aggregation and technology-enabled perishables supply chain and quoted Prakash on transforming that chain. Exact terms, ownership, channel economics, and board gates remained undisclosed. [investor-stated · moderate] Global AgInvesting

Decision freeze C — September 2022 capital decision. A March 2020 hospitality-demand collapse had forced a B2B-to-D2C pivot. By 2021, public accounts warned that produce was capital-intensive, two-city scale conflicted with national ambition, brand trust would take years, and logistics and affordability were hard constraints. [contemporaneous record · strong] Forbes India · YourStory

• A no-hindsight gate should require city-level contribution margin after spoilage/fulfillment, repeat cohorts, channel-specific margins, order density, inventory loss, cash conversion, farmer retention, runway without another round, and a reserve/stop rule. These are analyst-created controls. [researcher inference · strong] 2021 warning record

Reader exercise — decide before revealing the actual decisions or outcome. [researcher inference · strong]

1. Would you invest then? Make separate 2018, 2020, and 2022 decisions; do not let the pandemic excuse every post-pivot metric.
2. What would kill the deal? Set channel-level contribution, spoilage, density, repeat, working-capital, and runway thresholds.
3. What ownership/terms fit the risk? Specify milestone tranches, board information, reserve cap, liquidation terms, and a channel stop rule.
4. What evidence unlocks the next check? Require stable city cohorts, positive post-fulfillment contribution, controlled waste, farmer retention, and cash conversion.

Actual decisions and outcome. Accel joined the 2018 seed, 2020 $5.5 million Series A, and a 2022 $12.5 million round led by IvyCap with existing investors participating; the 2022 report also noted stress at comparable operator Fraazo. [retrospective witness · moderate] Accel page · 2020 round · 2022 round A commercial registry mirror associates Prakash with a nominee-director role, but an MCA DIR-12, minutes, votes, intervention record, and stop authorization were not recovered. [retrospective witness · provisional] Zauba record ROC-derived FY24 reporting shows ₹36.81 crore total revenue, ₹88.62 crore expense, ₹52.21 crore loss, and negative ₹36.55 crore operating cash flow; Entrackr later reported operations would stop after failed scaling/fundraising. A commercial registry mirror continued to show the legal entity as active, so this is a reported operational shutdown, not proved liquidation or zero recovery. [retrospective witness · moderate] TheKredible · Entrackr shutdown report · Zauba record No case-specific Prakash lesson was found in the inspected sources; his later climate-tech underdelivery admission is sector context, not proved causation. Whether staged financing, refusing the 2022 round, or an earlier channel stop improved recovery is unknown. [researcher inference · strong] Mint interview · search boundary

Portfolio-level forensics

What the observable portfolio actually says

• The mutable 28-company Accel graph contains 15 seed labels, eight Series A labels, and five year-only records with no stage. Initial-year fields cluster as 2010:1, 2011:1, 2012:2, 2014:2, 2015:5, 2017:4, 2018:1, 2019:3, 2020:1, 2021:1, 2022:1, 2023:3, 2024:1, and 2026:2. [researcher inference · strong] official graph and reciprocal pages
• Four current cards show acquisitions—CapriCoast/HomeLane, GlowRoad/Amazon, QwikCilver/Pine Labs, and Travelstop/Wego—and BlueStone has separate primary IPO evidence. These are five observable liquidity events among the current 28, not five wins or a 17.9% realized-return rate. [researcher inference · strong] official graph · BlueStone prospectus
• The 52-group discovery ledger adds removed, historical, personal, board, essay, and counter-attribution cases. It is deliberately heterogeneous and cannot support a statistical loss rate, stage mix, or sector-performance table. [researcher inference · strong] Denominator reconciliation
• Sector classifications are not consistently published across the 52; check size, security, ownership, dilution, reserve allocation, board activity, and cash flows are missing for most cases. Any sector or ownership concentration inferred from the visible names would reproduce profile-selection and survivorship bias. [researcher inference · strong] Comprehensive ledger
• Case-specific governance evidence is materially narrower than website association. Official filings support Prakash board/nominee roles at BlueStone, Infra.Market, and RentoMojo; HealthcareMagic and UrbanTouch supply only commercial-registry association leads, while UrbanTouch's legal record is vehicle-only. None publishes a complete sourcing-to-exit decision chain. [researcher inference · strong] material-case role matrix

Capital, follow-ons, and fund-performance controls

• The visible vehicle sequence rises from a reported $10 million Erasmic corpus to $60/$65 million Fund II, $155 million India III, a $305 million India IV offering before first sale, $430 million India V offering/$450 million reported close, $525 million India VI Form D sale/$550 million announcement, and $650 million India VII and VIII. These figures reflect close, offering, sold amount, or announcement scope and are not one continuous audited series. [researcher inference · strong] Vehicle table
• A Rockefeller Foundation 2019 filing carried India III, IV, V, and VI interests at approximately $13.72 million, $24.15 million, $10.41 million, and $0.46 million. These are year-end fair values—not commitments, paid-in capital, distributions, or returns. [contemporaneous record · strong] Form 990-PF, p. 31
• Fortune's 2018 account reported more than 35 exits, including 18 M&A outcomes, cited Tracxn estimates from 1x to 12x, and relayed Accel's statement that its “first fund” had returned capital while its “second” was expected to deliver attractive returns. The article does not identify the exact legal vehicles behind those labels or supply cash-flow dates, denominator, gross/net basis, or audit. [retrospective witness · moderate] Fortune India
• The same account described a 12–15 year horizon and more than 40 follow-on companies. That supports patient-capital and reserve-direction claims at firm level, not Prakash's allocation, reserve discipline, or realized performance. [retrospective witness · moderate] Fortune India
• Fortune also quotes a non-consensus process in which one partner could approve a small deal and two a somewhat larger one, but supplies no thresholds, veto rules, period/fund scope, or Prakash-specific authority. [retrospective witness · moderate] Fortune India, decision-process passage

Team outcomes that are not Prakash returns

• Swiggy's prospectus records India IV selling 10,572,706 shares at ₹390 for ₹4,123.36 million gross and a ₹11.17 weighted-average acquisition cost; the current Accel page names Anand Daniel, not Prakash. [contemporaneous record · strong] Swiggy prospectus · Accel page
• Freshworks disclosed a 25.5% pre-IPO Accel affiliate pool and a roughly $100 million December 2019 repurchase from an India III affiliate; Accel's official account is by Sameer Gandhi and Shekhar Kirani, and disclosed voting authority did not name Prakash. [contemporaneous record · strong] Freshworks prospectus · Accel account
• Accel's Flipkart account says the firm invested $800,000 but is authored by Gandhi and Subrata Mitra. The Padma citation's broad credit is retained as conflicting evidence, not used to transfer deal economics to Prakash. [contemporaneous record · strong] Accel account · Padma citation

Failure, miss, and anti-portfolio postmortems

Prakash later said Indian climate technology had not lived up to its potential and described a narrower advanced-manufacturing screen around defensible IP and engineering. This is a sector-level underdelivery admission and strategy refinement; it does not prove which deal caused the change or that later returns improved. [investor-stated · moderate] Mint interview

Uncertainty and retrieval ledger

Economics and ownership

• No audited Prakash-personal or fund-by-fund DPI, TVPI, IRR, gross/net bridge, loss ratio, capital-account history, carry allocation, management-fee schedule, GP commitment, hurdle, clawback, or cash-flow ledger was located. [researcher inference · insufficient public record] Vehicle evidence
• BlueStone lacks lot-level basis and full India III cash history; QwikCilver lacks consideration allocation; Deeksha and HealthcareMagic lack seller waterfalls; UrbanTouch lacks fees, taxes, timing, dilution history, and any unobserved distributions; RentoMojo's OFS price remains blank; Infra.Market and BookMyShow lack completed liquidity evidence. [researcher inference · strong] Economic ranking
• An official ITAT order concerns Prakash's sale of land and a villa, allows his appeal on separate long- and short-term capital-gains treatment, and states that the interest-deduction issue was not before the tribunal. It does not concern Mu Sigma, an investment exit, or venture economics; it is retained only as an identity-controlled legal record. [contemporaneous record · strong] ITAT order, paras. 2–7 and 14–18

Attribution, governance, and denominator

• The 52-group union cannot be converted into a career hit rate: profile removals/additions are unaudited, the Padma wording has no roster, the 15-company Erasmic total is incompletely named, and private or personal positions may be absent. [researcher inference · strong] Denominator table
• A DIN/company registry associates the matching identity with 71 legal entities, but that set mixes fund-management, property/family, nonprofit/civic, duplicate/alias, operating, and unresolved commercial entities. It is a governance lead list, not an investment denominator. [retrospective witness · provisional] Zauba DIN page
• IC thresholds, vetoes, sourcing, diligence, negotiation, allocation, follow-on approvals, board votes, recruiting interventions, exit influence, geographic authority, succession, and partner carry shares remain unknown except where a case source supplies a narrower role. [researcher inference · insufficient public record] People map
• Company failure, an Accel vehicle's exposure, Prakash's website/board association, and Prakash's personal economics are four separate fields. No adverse record found justifies collapsing them. [researcher inference · strong] Failure postmortems

Vehicle and chronology reconciliations

• Erasmic's legal close and allocations remain unavailable; Fund II's $60/$65 million, India V's $430/$450 million, and India VI's $525/$550 million pairs remain disclosure-scope differences without a public bridge. [researcher inference · strong] Vehicle table
• India VII's entrepreneur sleeve should not be mechanically added to the $650 million headline; India IV Form D related-person listings and later Form D omissions are not complete IC, ownership, or carry rosters. [researcher inference · strong] India VII Form D · India VIII Form D
• RentoMojo's 2015 board-join announcement and its DRHP's 4 December 2024 nominee-director date need the underlying appointment/cessation filings to reconcile tenure. [researcher inference · strong] RentoMojo DRHP

Primary records that could change the assessment

• Obtain Accel/Erasmic allocation schedules, IC memos, LP reports, capital accounts, side letters, cash-flow schedules, carry agreements, and reserve models. [researcher inference · strong]
• Obtain MCA cap tables, PAS-3/MGT-7/SH-7/DIR-12 filings and board minutes for the material current and removed cases; obtain seller waterfalls and closing statements for BlueStone, QwikCilver, Deeksha, HealthcareMagic, UrbanTouch, BabyOye, GlowRoad, Travelstop, CapriCoast, and Exclusively.in. [researcher inference · strong]
• Obtain company confirmation of Deep Rooted's operating/legal status, Good Glamm lender recoveries, and Prakash/Accel write-offs or recoveries; commercial registry mirrors remain leads until underlying filings are retrieved. [researcher inference · strong]
• Retrieve a roster defining the Padma citation's company and board counts, historical Accel profile revisions, and the ten unnamed portions of Erasmic's reported 15-company portfolio. [researcher inference · strong]

Closing search and adverse boundary

The final three case-specific searches—UrbanTouch + Prashanth Prakash + consideration + Accel, HealthcareMagic + Prashanth Prakash + earnout + Accel, and Deeksha + Prashanth Prakash + proceeds + Accel + Vedantu—produced no additional ownership, IC, or distribution evidence beyond the records cited above. [researcher inference · strong]

Identity-controlled searches combined Prakash with loss, write-off, fraud, sanction, SEBI, NGT, Supreme Court, Deep Rooted, Good Glamm, and exact company variants. They found corporate failures and a personal land-and-villa capital-gains appeal, but no investment-related fraud or securities-sanction finding against Prakash. The ITAT allowed his appeal on separate capital-gains treatment and did not decide the interest-deduction issue; the order supplies no investment-return evidence. Discovery Village litigation was dismissed by the NGT and the Supreme Court declined to interfere; allegations are not merits findings. Search absence is not proof that no private, sealed, or inaccessible matter exists. [researcher inference · strong] ITAT order, paras. 2–7 and 14–18 · NGT reproduction · Supreme Court order

Bottom line

The most defensible public conclusion is narrower than the reputation: Prakash helped build an early India seed institution, has a long and specifically documented BlueStone relationship, and appears in several valuable or instructive transactions. Yet the record also contains likely write-offs, reported shutdowns, distressed roll-ups, team-level misses, ordinary companies, and many attribution conflicts. Without the private denominator and cash-flow ledger, this is a decision-forensics curriculum—not a verified personal performance table. [researcher inference · strong] 52-case ledger · Portfolio forensics

Coverage and evidence base

Literal priority-source reconciliation

[contemporaneous record · strong] The source map now contains 26 priority written records: 21 ★★★ and 5 ★★. WRITINGS added two reports, three sole-bylined op-eds, one Prakash-bylined civic essay with TNN page credit, a signed foreword, two stable direct LinkedIn posts, and nine contemporaneous Erasmic pages absent from the original SCOUT denominator. The reports remain joint institutional works, and the Erasmic pages remain unbylined team/vehicle voice rather than sole-authored Prakash doctrine. The public-access denominator is 24 full + 2 partial = 26.

[contemporaneous record · strong] for the inspected bylines, page credits, signatures, institutional authorship, and access states; [researcher inference · strong] for the resulting attribution limits. The corpus has six attribution classes. The BlueStone reflection and five policy/manufacturing op-eds are clearly solely bylined to Prakash; the civic museum essay is Prakash-bylined with TNN page credit and a footer identifying Prakash as writer. Two LinkedIn posts are direct personal-channel writing, and the 2024 Bengaluru report contains a separately signed Prakash foreword. Captain Fresh, MyGlamm, Golden Age, omnichannel, and advanced-manufacturing are joint works whose arguments show public endorsement, not sentence-level authorship. The live Accel profile is attributed first-person institutional copy but mutable and undated. Unboxing Bengaluru is co-authored with Malini Goyal and only partially accessible. The nine Erasmic pages are contemporaneous unbylined institutional records: strong evidence of what the small founding team represented publicly in 2008, but not personal authorship or unilateral practice. Accel profile · Moneycontrol author index · Times of India author index · Penguin book page · Common Crawl Erasmic index

Evidence boundary — [researcher inference · strong]. This is a broad but uneven public corpus spanning 2008–2026: unusually rich in early institutional doctrine and recent thematic writing, thin in the intervening years. It supports dated positions on founder selection, locally specific markets, technology-enabled supply chains, consumer brands, omnichannel scaling, manufacturing, and patient company building. It does not reveal a complete declined-opportunity set, contemporaneous IC memoranda, ownership, reserves, follow-on decisions, write-offs, fund cash flows, or audited personal returns. Portfolio essays written after positive milestones are selected retrospective narratives; co-authored firm reports are interested advocacy. No repetition count converts those limitations into efficacy evidence.

The search included exact-name and name-variant sweeps across Accel, SeedToScale, Medium, Economic Times, Moneycontrol, old Erasmic/Accel India domains, LinkedIn/X, policy and conference publications, reports, Google Books, publisher pages, library catalogs, Internet Archive, and Common Crawl. Wayback CDX routes timed out, but Common Crawl's 2008–09 index and ARC records recovered nine full original Erasmic pages with capture timestamps, offsets, lengths, and digests preserved in the notes. Two full-body LinkedIn posts were identity-verified; no stable complete social corpus or attributable policy submission was found. Search results for a different Prashanth Prakash in engineering/medical contexts were excluded by identity controls.

Source-by-source analysis

Erasmic — Funding Philosophy (captured 2008; institutional voice)

Published screen. Erasmic wanted most, not necessarily all, of six traits: roughly 6–12 months of operation and customer validation; differentiated services made harder to copy by technology or scarce knowledge; comparatively low capital needs; a credible India revenue path; a domain where the managers could add more than money; and working compatibility with the founding team. Wireless/telecom and consumer web were named areas despite the claim that the market was too small for strict sector bias. [contemporaneous record · strong] for the firm's screen; [contemporaneous record · provisional] for asserted capital efficiency and valuation effects. Common Crawl index record · source note

Use and limits. The historical rule is evidence before financing, differentiated delivery, manageable capital, manager-founder fit, and a path to sustaining revenue. [researcher inference · strong] It is among the closest public records to a period selection screen, but the capture does not establish original publication or revision history; it remains unbylined team copy and shows no applicant funnel, scoring, exceptions, valuation, or outcome comparison. It cannot be assigned personally to Prakash or treated as proof that the screen worked. Funding Philosophy, six paragraphs and archive locator

Erasmic — Investment Strategy (captured 2008; institutional voice)

Argument and causal chain. The page prioritizes differentiated services where technology or scarce skills create a replication barrier, an India operation reinforces the advantage, capital consumption stays modest, and a strategic acquirer can recognize the value. India is framed as an operating advantage within a cross-border company, not only an end market. [contemporaneous record · strong] for the published chain; [contemporaneous record · provisional] for its economic validity. Common Crawl index record · source note

Decision implications and limits. This is the corpus's earliest explicit liquidity logic: build scarce capability for strategic acquisition. A modern translation should require named buyer classes and independently valuable economics because an acquisition-dependent thesis can fail even when the product advantage is real. [researcher inference · strong] The page provides no buyer map, integration risk, security, ownership target, price discipline, or holding-period rule. Investment Strategy, complete body and archive locator

Erasmic — Our Value Add (captured 2008; institutional voice)

Operating model. Erasmic promises hands-on support across strategy, customers, organization, technology, operations, marketing, and finance: secure strategic customers or development partners; recruit managers; develop IP and technical staff; translate customer requirements; and provide controller/CFO help for operations, fundraising, and debt. [contemporaneous record · strong] for the promised service model, not its capacity or efficacy. Common Crawl index record · source note

Decision implications and limits. The implied method is to diagnose a missing capability, add expertise or access, install a repeatable process, and shift founders toward roles only they can fill. The breadth creates a capacity question the page never answers: four-to-six intensive companies may be consistent with this promise, but frequency, staff allocation, transfer milestones, founder consent, and outcomes are absent. [researcher inference · strong] It is a claimed institutional model, not evidence that Prakash personally performed every function. Our Value Add, six functions and archive locator

Erasmic — Overview (captured 2008; institutional voice)

Fund design and gap. The page proposes—not proves a legal close of—a US$10 million fund for cross-border US/India startups using India delivery and US sales. It describes larger foreign funds as preferring more than US$2 million while local companies needed smaller cheques plus mentoring, then proposes four to six companies, US$100,000–US$500,000 each, and substantial senior-management time. [contemporaneous record · strong] for intended design; [contemporaneous record · moderate] for the unsampled capital-gap claim. Common Crawl index record · source note

Portfolio implications and limits. The useful historical rule is to enter below large-fund cheque minimums only where concentration permits operating work that changes financing readiness. Four-to-six initial cheques at the disclosed endpoints account for only US$0.4–3 million, so the page does not reconcile the rest of the proposed corpus; reserves, follow-ons, fees, uncalled capital, or a broader portfolio must not be invented. [researcher inference · strong] Later contemporary reporting is needed to distinguish this proposal from the fund that actually closed. Overview, proposal language and arithmetic boundary · Mint, November 2008 fund close

Erasmic — Indian Landscape (captured 2008; institutional voice)

Market map and capital gap. The page treats India in three distinct roles: a large emerging customer market; an operating base for KPO/BPO and cross-border delivery; and a lower-cost validation environment before entering more expensive geographies. Mobile/television reach, telecom spending, broadband, disposable income, cricket, and Bollywood support its domestic-demand story, while it argues that funds above US$100 million generally could not make sub-US$1 million investments. [contemporaneous record · strong] for the dated firm thesis; [contemporaneous record · provisional] for unsourced adoption counts and the claimed capital pool. Common Crawl index record · source note

Decision implications and controls. The historically faithful rule is to identify the mismatch between company needs and incumbent-fund cheque sizes, then state whether India is indispensable as customer market, operating base, or validation environment. China-company analogies establish a search direction, not comparable regulation, income, infrastructure, or outcome. [researcher inference · strong] Period reporting independently supports a small-cheque gap but not every aggregate or the page's precise forecast. [contemporaneous record · moderate] Indian Landscape, three sections and period controls

Erasmic — Founding Team (captured 2008; institutional voice)

Team and biography record. Erasmic presented four general partners—Subrata Mitra, Prashanth Prakash, Mahendran Balachandran, and Gagan Kumar—and emphasized founder/operator experience, cross-border technology work, large-company management, and engineering/business education. Its Prakash biography traces S3 Technologies, US consulting, the 1998 Bengaluru establishment of NetKraft, its 2004 Adea sale, prior venture fundraising, and technical education. [contemporaneous record · strong] for what the firm represented; [contemporaneous record · moderate] for biography claims, some of which period transaction reporting corroborates. Common Crawl index record · Rediff/PTI NetKraft transaction · source note

Attribution and limits. The page says Mitra and Prakash co-founded an Erasmic Consulting precursor before the 2006 fund and invested personal time and capital, but it does not define the entities, dates, individual economics, authority, allocation, or IC rights. NetKraft headcount differs between the archived bio and contemporaneous press. The defensible people signal is partnership-level experiential complementarity—not proof that the team selected founders in its own image or that operating experience caused returns. [researcher inference · moderate] Founding Team, four biographies and attribution controls · Rediff/PTI, NetKraft transaction

Erasmic — Investment Area (captured 2008; institutional voice)

Eligibility lanes. The page names three cross-border cases: an Indian company with some customer validation and usually early revenue but little non-Indian market access; a US founder needing a strong India development/BPO back end; or a US founder whose idea could attract later venture capital after Erasmic helped form a team, prototype, and early validation. [contemporaneous record · strong] for eligibility; [researcher inference · moderate] for the shared bridge logic. Common Crawl index record · source note

Decision implications and limits. The relevant test is whether cross-border market access or execution is actually the binding constraint and which milestone should unlock institutional financing. Thresholds such as “some” traction remain discretionary, and the second lane has no explicit validation gate. [researcher inference · moderate] Entity, sector exclusions, valuation, ownership, reserves, authority, board expectations, follow-ons, and exits are absent; “considered for funding” is a funnel, not proof of investment. Investment Area, three eligibility bullets and archive locator

Erasmic — Track Record (captured 2008; institutional voice)

Intervention ledger. The page lists Galvanon, Small Device Technologies, Mu Sigma, NuWare, and Pension Retirement Online, but does not consistently distinguish investment, client, incubation, personal angel position, or fund asset. Mu Sigma and PRO contain the substantive evidence: Erasmic collectively claims seed-network help, interim operations, India team formation, delivery/process work, automation, presales, and US commercial access. [contemporaneous record · moderate] for self-reported team actions; [contemporaneous record · strong] for the ambiguity of the category labels. Common Crawl index record · source note

Controls and limits. SEC and trade records support the Galvanon and Small Device acquisitions but do not connect Prakash or Erasmic economics; the Small Device month conflicts. [contemporaneous record · moderate] NCR 2005 Form 10-K, acquisition note · RCR Wireless, Small Device transaction NuWare has only a business description. No page entry supplies complete dates, individual owners of interventions, amounts, governance rights, causal outcomes, or proceeds. The responsible interpretation is a favorable institutional activity ledger requiring company, board, vehicle, and cash-flow controls—not a personal or audited return record. [researcher inference · strong] Track Record, five entries and archive locator

Erasmic — Positioning (captured 2008; institutional voice)

Claimed differentiator. Erasmic calls itself a multi-skilled, hands-on partner and says partners would contribute engineering, delivery-process, sales/marketing, business-development, and general-management skills. [contemporaneous record · strong] for positioning; [contemporaneous record · provisional] for the promotional suggestion that engagement could ensure success. Common Crawl index record · source note

Underwriting the service. Support should be treated as a scarce resource: name the responsible partner, time commitment, concrete deliverable, founder consent, handoff condition, and opportunity cost. Those safeguards are analyst-created; the page gives no hours, authority, duration, comparator, adverse case, or portfolio-capacity data. [researcher inference · moderate] It establishes intended differentiation, not uniqueness or efficacy. Positioning, sole body paragraph and archive locator

Accel profile and current first-person copy (mutable; accessed 2026-08-19)

Context, argument, and reasoning. The profile is not a dated investment essay. It combines an institutional biography with short first-person answers: early simulator-interface work taught Prakash to connect engineering with user needs; he describes productivity as aligning work to energy and converting long-form media to transcripts; and he adopts a collective-credit maxim. The defensible reasoning chain is experience building interfaces → attention to what users need → early product-management orientation. [investor-stated · strong] for these self-descriptions; [researcher inference · provisional] for treating them as a historical investment screen. Accel profile, “What was your first job?” and advice answer · source note

Decision implications, assumptions, and limits. The investable translation is to examine whether a technical founder understands both implementation and user behavior, and whether learning systems reduce search cost without replacing verification. [researcher inference · moderate] The credit answer is evidence of a preferred working norm, not proof that team attribution was always handled correctly. The page says he joined Accel in 2008 and currently focuses on consumer, marketplaces, and SaaS; its relationship cards are a mutable discovery graph, not a career ledger or personal ownership record. It supplies no examples of rejected investments, pricing, portfolio construction, reserves, governance, or exits. Accel profile, About and Relationships

“The Golden Age of Indian Startups, VCs” (2020; with Mahendran Balachandran and Subrata Mitra)

Context, argument, and reasoning. The three authors look back over Accel/Erasmic's first 15 years in India and argue that domestic consumption, a young population, expanding internet access, localized solutions, technology-led supply-chain aggregation, and globally competitive SaaS created a compounding startup opportunity. Their chain is macro adoption and local demand → founders solving India-specific access and coordination failures → digitally enabled scale → ecosystem reinvestment and export opportunity. [investor-stated · strong] for the jointly stated 2020 position; [retrospective witness · moderate] for the prior firm history; [investor-stated · provisional] for broad value-creation and returned-capital claims because the article provides no audited denominator. Medium essay, opening through four themes · source note

Evidence, implications, and limits. The lunchtime Flipkart order-density anecdote and the comparison between Indian domestic consumption and export-led peer economies make the local-market mechanism concrete. Later sections identify consumer brands, B2B aggregation, and SaaS export as distinct lanes rather than one generic “India growth” thesis. [investor-stated · strong] The useful decision rule is to specify which adoption bottleneck a company removes and why a local founder can translate a global model rather than merely copy it. [researcher inference · strong] Yet the article groups Accel's wins, uses market capitalization and valuations alongside economic impact, and says venture money returned “in multiples” without vehicle-level cash flows. It is a firm narrative, not Prakash-personal credit or return evidence. Medium essay, paragraphs on internet access and SaaS · source note

“Transforming the Seafood Supply Chain” (2021; with Barath Shankar Subramanian and Rachit Parekh)

Context, argument, and reasoning. Published with Accel's Series A announcement, this is the closest accessible near-decision, post-decision public rationale—not a reproduced ex-ante memo. The authors start from a fragmented, heterogeneous, perishable supply system: uncertain catch by species/size/quality, weak cold chain, buyer-specific demand, and constrained global supply. They argue that aggregation, prediction, traceability, end-to-end logistics, and demand matching can improve fill rates and reduce spoilage. [investor-stated · strong] for the jointly announced rationale; [investor-stated · provisional] for the uncited $50 billion-plus India and $100 billion-plus export market figures. Captain Fresh essay, opening through supply-demand matrix · source note

Founder and decision evidence. The founder screen is lived domain obsession plus the ability to teach the investor a market the investor initially understood only vaguely; the team says it tracked Utham Gowda for a year before investing. [investor-stated · strong] The implied rule is not simply to prefer “under-the-radar” sectors, but to require a founder-specific model for converting fragmentation into standardized service. [researcher inference · strong] The page cites other Accel B2B companies as analogies, yet gives no pricing, security, ownership, margin, working-capital, quality-loss, regulatory, or follow-on evidence. Joint authorship prevents Prakash-only attribution; Accel's current company card names Barath Shankar Subramanian as the associated partner, adding a separate control against personal deal credit. [contemporaneous record · strong] Later company outcomes cannot be backfilled into this 2021 thesis. Captain Fresh essay, first-meeting and approach passages · Accel company card · source note

“Building a digital CPG giant in India: Our investment in MyGlamm” (2021; with Anand Daniel and Rachit Parekh)

Context, argument, and reasoning. At a growth investment, the authors argue that informed consumers, dissatisfaction with standardized incumbent products, digital discovery, content, direct channels, and rapid feedback create room for a new beauty/CPG platform. Their company thesis rests on a founder willing to own distribution, use multiple brands for distinct segments, and combine POPxo's community with commerce. [investor-stated · strong] for the jointly stated 2021 rationale; [investor-stated · provisional] for the “generational company” forecast and content-to-commerce flywheel because no cohort economics or comparison is published. MyGlamm essay, consumer shift through investment announcement · source note

Counterevidence and implications. The founder screen emphasizes ambition, decisiveness, resilience after near-failure, and opportunistic recombination during COVID. [investor-stated · strong] That yields a practical question set—what hard distribution choice has the founder made, what customer insight is proprietary, and does content lower acquisition cost without obscuring retention or working capital? [researcher inference · strong] Attribution is internally uneven: the page displays three authors, while its closing sign-off adds Manasi beyond those displayed authors, so it cannot support sentence-level or Prakash-personal deal credit. [contemporaneous record · strong] The later Good Glamm breakup is a necessary control: reported flat financing, defaults/arbitration, delayed salaries, asset sales, and an Accel board resignation show that a persuasive flywheel narrative did not eliminate financing, governance, or integration risk. This does not establish Prakash's individual loss or responsibility, but it falsifies any unqualified success claim. [researcher inference · strong] MyGlamm essay, displayed authors and closing sign-off · Economic Times, reported collapse timeline · source note

“Innovation, sustainability to fuel transformative phase” (2023)

Context, argument, and reasoning. This sole-authored year-ahead op-ed predicts that AI, manufacturing-as-a-service, renewables, EVs, engineering talent, and digitally disaggregated supply chains will move India from a service-centric economy toward products and manufacturing. It connects a large developer base to Western software/B2B demand, AI to access in health and education, and China-plus-one style supply shifts to the Global South. [investor-stated · strong] for the dated forecast; [investor-stated · provisional] for its market sizes and causal confidence because methodology and falsification criteria are absent. Economic Times op-ed, body paragraphs 131–137 · source note

Decision implications and limits. The article is a theme map, not an underwriting process. To make it decision-useful, an investor would need to identify the enabling policy or cost curve, the product layer that captures value, the customer's switching incentive, and a dated indicator that would disconfirm the theme. [researcher inference · strong] Its optimism also sits beside Prakash's June 2023 statement that India lacked a real deep-tech research and financing ecosystem; the corpus therefore records aspiration and constraint in the same year rather than treating the forecast as settled fact. [researcher inference · strong] Moneycontrol, 2023 deep-tech critique · source note

“Decoding Omnichannel: Strategies for D2C Brands” (2024; six authors)

Context, argument, and reasoning. The report rejects online-versus-offline as a binary. It proposes a staged loop: verify online awareness and local order density; seed a base city; deepen the city and test new geographies/formats; identify look-alike markets and cannibalization; then scale using data and ML. Lenskart, BlueStone, and AgroStar illustrate catchment selection, touch-and-try, cross-channel inventory, content-to-store conversion, and franchise experiments. [investor-stated · moderate] because Prakash is one of six authors; [researcher inference · provisional] for efficacy because only selected cases and company-supplied metrics are visible. Report page, framework and case studies · source note

Decision implications and limits. This is the corpus's most operational scaling framework. [researcher inference · moderate] A board can ask for awareness, order-density, drive-time, footfall, contribution-margin, inventory-turn, cannibalization, and repeat-purchase evidence at each gate. It extends MyGlamm's digital-first thesis and BlueStone's trust thesis by treating stores as measured complements to online discovery. [researcher inference · moderate] The page does not link the forecast model, raw case data, failed rollout cohort, or stable separate report download. The correct use is a test sequence, not the assertion that every D2C brand needs stores or that reported growth was caused by omnichannel expansion. Report page, framework and case studies · source note

“India’s manufacturing journey is set to get better in 2025” (2024)

Context, argument, and reasoning. The sole-authored op-ed sharpens the prior year's manufacturing optimism into four pillars—precision engineering, R&D-led manufacturing, robotics, and defence—plus an AI-human collaboration thesis. It points to engineering talent, China-plus-one, governance continuity, corporate manufacturing commitments, policy support, public examples, and falling robotics costs. [investor-stated · strong] for the dated taxonomy; [investor-stated · provisional] for the many market, deal, talent, export, cost, and adoption figures because underlying datasets are not linked. Moneycontrol op-ed, paragraphs 320–369 · source note

Assumptions, implications, and contradictions. The implied screen favors globally benchmarkable technical output, owned R&D, automation serving an actual labor/quality constraint, and policy-supported strategic demand. [researcher inference · strong] The AI-human section rejects pure substitution but offers no task-level productivity, error, adoption, or margin evidence. Company examples include Accel-linked businesses without disclosing selection or economics, so they are illustrations rather than performance validation. [researcher inference · strong] This piece is more specific than the 2023 forecast but still lacks capitalization, certification, sales-cycle, and reserve rules essential for venture underwriting. Moneycontrol op-ed, four pillars and AI-human section · source note

“Celebrating BlueStone’s IPO” (2025)

Context, argument, and reasoning. This sole-authored post-IPO reflection says Accel mapped e-commerce in 2011 with a “prepared mind,” identified jewelry as a design/brand/supply category rather than a marketplace-placement or price game, and backed founder Gaurav Singh Kushwaha's locally specific view that Indian demand would center gold, design, and everyday wear. The operating chain is trust problem → high-fidelity rendering, home try-on, and exchange policy → integrated design/manufacturing/supply → measured omnichannel distribution. [investor-stated · strong] for Prakash's retrospective thesis; [documented behavior · strong] for the long board relationship; [researcher inference · provisional] for claiming each insight existed fully at entry. BlueStone essay, “Why we leaned in” through omnichannel · BlueStone annual report, board and security sections · source note

Implications and limits. This is the strongest Prakash-specific deal writing because it combines category selection, founder assessment, product mechanisms, capital-cycle endurance, and a public milestone. [researcher inference · strong] It suggests three rules: local behavior can invalidate a global category template; trust may require full-stack capabilities; and a capital-scarce middle rewards frugality and unit-economics discipline. [researcher inference · strong] It does not disclose the original memo, terms, ownership, follow-ons, dilution, cash proceeds, missed milestones, or board interventions. The IPO is not identical to a realized personal or fund return, and retrospective selection after listing remains an important bias. BlueStone essay, thesis and endurance sections · BlueStone annual report, board/security controls · source note

“India’s Advanced Manufacturing Leap” (2025; with Prayank Swaroop, Vidhya Ananthakrishnan, and Nayan Nandan)

Context, argument, and reasoning. The report defines three archetypes: innovation in the product; proprietary equipment or raw material that changes a process; and process innovation using largely standard equipment. Its qualification screen asks for automation/digitization, owned IP or proprietary integration, and clear performance or cost gains. The broader thesis is that supply-chain diversification, engineering talent, public incentives, and domestic demand can support high-precision, IP-led Indian companies in electronics, aerospace/defence, and automobiles. [investor-stated · strong] for the jointly published taxonomy; [investor-stated · provisional] for the macro forecast and promotional comparisons. Rendered report, definition and qualifying framework · source note

Decision implications and limits. This is more decision-ready than the two op-eds: identify the innovation archetype, verify IP and integration ownership, benchmark performance, then map qualification, capex, working capital, supplier/customer concentration, policy dependence, and commercialization time. [researcher inference · moderate] The report acknowledges a remaining quality gap but does not publish failure rates, certification delays, capital overruns, ownership requirements, or venture-return base rates. Its disclaimer says Accel/Praxis/public data may be incomplete and forecasts may differ from outcomes. Public-company successes and selected startups therefore establish plausibility, not a portfolio result. Rendered report, framework and disclaimer · source note

Unboxing Bengaluru: The City of New Beginnings (2022/2023; with Malini Goyal; partial)

Accessible argument and attribution. Publisher and library records identify Goyal and Prakash as co-authors, while authorized Penguin, Moneycontrol, and Times of India excerpts expose limited book text. The ecosystem excerpt argues that compact geography, technology-centered culture, educated migration, porous professional networks, and pay-it-forward behavior amplify repeated founder/operator encounters. The social excerpt records founders' illiquid paper wealth, long hours, unstable early years, and limited relationship bandwidth. [investor-stated · moderate] for the joint accessible thesis; [retrospective witness · moderate] for quoted founder experiences. Neither sentence-level authorship nor the unopened chapters may be assigned to Prakash. Penguin book page · authorized Moneycontrol excerpt · authorized Penguin excerpt · source note

Decision implications and limits. The defensible translation is to test an ecosystem for talent density, repeated local interaction, accessible role models, cultural openness, and civic capacity rather than treating capital as the only input. [researcher inference · moderate] The Ather anecdote concerns reported Sachin and Binny Bansal financing, not Prakash or an Accel vehicle. Anecdotes do not establish comparison-city base rates, network inclusion, causality, or returns. Edition dates and page counts conflict, and the complete book, notes, bibliography, and “Fault Lines and Futures” treatment still require licensed access. Authorized ecosystem excerpt · source note, edition and access ledger

“A museum in Bengaluru will showcase India’s technology, innovation & startup feats” (2022)

Context, argument, and evidence. This Prakash-bylined civic essay—with TNN page credit and a footer identifying Prakash as writer—proposes a living museum that preserves India's technology and startup history, interprets the present, and continually refreshes possible futures. Its ecosystem map treats government laboratories, academia and public enterprises; domestic and multinational corporations; and startups as coequal contributors. [investor-stated · strong] for Prakash's proposed institution; [researcher inference · moderate] for translating its network view into investing. A NASSCOM executive summary supports the essay's rounded FY2022 industry-revenue scale, but that forecast is not a venture outcome. [contemporaneous record · strong] Times of India essay · NASSCOM Strategic Review 2022, executive summary p. 8 · source note

Implementation control and limits. The public-private proposal later became the Karnataka Technology Innovation Museum Foundation, with Prakash as chair, so the essay has a documented institutional follow-through. Its original 18–24 month expectation did not hold: a 2026 official RFP still described Phase 1 and targeted later phases for 2027–28. [documented behavior · strong] That persistence-plus-delay is more informative than repeating the launch vision. The piece concerns civic memory and inspiration, not company selection, ownership, or returns. [researcher inference · strong] MIST project and board · MIST RFP, dated roadmap

“Budget 2023: Focus on emerging sectors praiseworthy, but startups need a helping hand” (2023)

Argument and primary-source controls. Prakash's two-part judgment is that infrastructure and emerging-sector programs can expand technology formation, while tax and regulatory rules still impede startup risk capital. He connects capex, agritech, green hydrogen, battery manufacturing, and AI centers to future demand, then criticizes non-resident angel-tax expansion and restrictive startup-exemption conditions while favoring GIFT IFSC simplification and capital-gains parity. [investor-stated · moderate] for the interested policy thesis. The Budget speech and Finance Bill confirm the announcements and proposed legal changes, but allocations, targets, and compliance counts are inputs—not realized employment, deployment, formation, or returns. Moneycontrol essay · Budget speech · source note

Decision implications and conflicts. The reusable rule is to map each policy to its legal instrument, beneficiary conditions, implementation date, transmission channel, and measurable company bottleneck. [researcher inference · strong] The article rarely supplies multipliers, take-up, counterfactuals, or capital-supply elasticity. Prakash's position as a venture investor makes his tax and fund-structure recommendations relevant but interested; the essay does not quantify who benefits or compare alternative public uses. [researcher inference · moderate] Moneycontrol essay, policy and startup-tax sections · source note, official controls

“Bullish on India: Watch out for startups bringing innovation and disruption to manufacturing” (2023)

Context, mechanism, and evidence. This sole-authored essay makes the 2023 manufacturing thesis explicit: demographics, digital skills, credit, policy, China-plus-one demand, multinational assembly, PLI incentives, digital public infrastructure, and physical infrastructure can support technology-enabled B2B manufacturing platforms. The company mechanism is coordination—aggregate fragmented suppliers, standardize processes, expose capacity and quality, match demand, and use data for procurement and monitoring. [investor-stated · moderate] for the thesis; [investor-stated · provisional] for market forecasts and selected company metrics. Moneycontrol essay · source note

Decision implications and counterevidence. A testable screen is to locate a fragmented, low-trust supply network, verify whether software improves capacity, price, quality or delivery visibility, then measure repeat demand, take rate, working capital, bad debt, quality liability, and owned-versus-leased capacity. [researcher inference · strong] The article omits those economics. Its Zetwerk, Infra.Market, and funding figures are company/report claims without complete definitions; later reporting of Infra.Market's owned plants complicates the “without fixed assets” characterization. [researcher inference · moderate] Inc42, later Infra.Market owned footprint The 2025 shift toward proprietary product/process/equipment and admission of earlier sector misses may be an underwriting refinement, but no decision ledger proves the change. [researcher inference · moderate] Mint, 2025 strategy reassessment · source note

“Seizing global opportunities: Budget 2024’s vision for India’s future” (2024)

Argument and accuracy controls. Prakash presents a competitiveness stack: talent, research finance, manufacturing infrastructure, quality, trade access, and capital-market reform should help India capture global opportunity. He links rural roads and agriculture spending to agritech, patient finance to deep-tech commercialization, vocational programs to workforce supply, industrial parks to export services, and credit/tax reform to private capital. [investor-stated · strong] for the policy framework; [investor-stated · provisional] for company outcomes. Official records expose two important errors or elisions: the proposed 10,000 facilities were bio-input resource centers, not general bio-research centers, and the ₹1 lakh crore item was financing capacity, not an additional annual R&D expenditure. Moneycontrol essay · Budget speech · source note

Decision implications and limits. The strongest rule is the essay's own access caveat: a headline pool matters only if implementation instruments, eligibility, fees, timing, co-financing, and operating access reach the target company. [researcher inference · strong] It supplies no causal evidence that announced programs increase productivity or venture returns. As with the 2023 piece, Prakash's public-policy and investor roles should be read together, with benefits, potential or perceived conflicts, disclosure/recusal controls, and alternative beneficiaries kept visible. The inspected public record is insufficient to establish influence, impropriety, or an actual conflict. [researcher inference · moderate] Moneycontrol essay, R&D-access caveat and capital sections · source note, official-speech controls

Bengaluru Innovation Report 2024 — signed foreword (2024)

Context, argument, and boundary. In a separately signed foreword, Prakash argues that digital technology broadens startup reach into traditional sectors and that Bengaluru's technology history, R&D centers, GCC/software talent, startup density, capital, accelerators, and public policy reinforce one another. The stated goal is comparative evidence and coordination toward a ten-year city vision. [investor-stated · strong] for the foreword; [researcher inference · moderate] for an investment-cluster screen. His Startup Vision Group chair and Accel partner titles make the public-private boundary explicit. Report PDF, foreword p. 5 · source note

Evidence and limits. The foreword's startup, unicorn, engineer, and IT-export figures are not footnoted there, and the 66-page report body has separate institutional/data authorship. The document can support a comparative cluster checklist—talent depth, incumbent R&D, formation, capital, infrastructure, and market access—but supplies no weights, causal identification, company-selection thresholds, or conflict-governance process. [researcher inference · moderate] No report-body claim is attributed to Prakash merely because he signed the foreword. [researcher inference · strong] Report PDF, signed foreword p. 5 · source note

“India’s next leap in global relevance…” — LinkedIn (2025)

Direct thesis and timing. The public activity ID fixes the post's creation at 2025-07-17, but LinkedIn marks the currently visible body “Edited” and exposes no edit history. The current wording says India's global relevance must extend from software and capital to IP-led manufacturing in defence, aerospace, semiconductors, renewables, and precision engineering; proprietary innovation, domain depth, regulated technical execution, and durable capability matter more than speed alone. [investor-stated · strong] for the body accessed 2026-08-19; [contemporaneous record · provisional] for assigning that exact wording to July 2025. LinkedIn post, body and Edited marker · source note, activity-ID derivation and edit boundary

Decision implications and limits. A practical screen is to identify a strategic technical bottleneck, test owned IP and domain knowledge, map regulation and qualification, and decide whether constraints compound defensibility or merely delay sales. That screen is [researcher inference · moderate]; the short post gives no stage, cheque, ownership, technical threshold, capital plan, or named decision. The creation date precedes Accel's August multi-author report, but whether the current exact wording anticipated it is [researcher inference · provisional] because the edit history is unavailable. LinkedIn post, current body and Edited marker · advanced-manufacturing report

National Startup Day reflection — LinkedIn (2026)

Latest located direct evolution. In a full public post dated by its activity ID to 2026-01-16, Prakash argues that India's advantage is shifting from speed alone to depth: global-standard products grounded in IP-led engineering and manufacturing. The goal is enduring, respected companies rather than a larger startup count. The linked PTI interview adds translational innovation, domain-immersed entrepreneurship, and global IP plus Indian manufacturing discipline, but those expansions are reported quotations rather than authored post text. [investor-stated · strong] for the post; [investor-stated · moderate] for the interview. LinkedIn post · The Week, full PTI syndication · source note

Implications and limits. Relative to the 2020 scale narrative, the post adds product endurance, owned IP, engineering depth, and manufacturing. A usable test is whether applied R&D or domain experience becomes difficult-to-replicate capability, whether production meets global standards, and whether speed accelerates learning without bypassing qualification. This is textual evolution, not proof that Accel's portfolio practice, allocation, or returns changed. [researcher inference · strong] 2020 Golden Age essay · 2026 LinkedIn post

Cross-source synthesis

Recurring beliefs and the evidence behind them

What the corpus does and does not support across the venture lifecycle

Intellectual chronology, evolution, and contradictions

Six tensions must remain visible:

1. Intuition versus instrumentation. An attributed 2020 joint interview/article records Prakash's guarded reliance on instinct and conditioning, while the omnichannel report insists on brand tracking, catchments, and staged data. The narrow reconciliation is intuition for where to look, followed by measurement before scaling; the corpus never states that complete rule. [researcher inference · moderate] YourStory joint interview/article · Omnichannel report
2. Global playbooks versus local first principles. BlueStone rejects a solitaire/global template, while the manufacturing essays depend on global supply-chain shifts and export demand. The consistent element is not local isolation but translating external opportunity through India-specific customer and capability evidence. [researcher inference · strong] BlueStone · Advanced Manufacturing
3. Digital-first efficiency versus full-stack control. MyGlamm celebrates direct/content channels; Captain Fresh and BlueStone require logistics, manufacturing, or stores. The choice appears contingent on the trust, quality, and coordination bottleneck—not a universal asset-light or vertically integrated preference. [researcher inference · strong] MyGlamm · Captain Fresh · BlueStone
4. Deep-tech deficit versus manufacturing confidence. A 2023 critique says the research/financing ecosystem is inadequate; the 2025 report says academia, talent, policy, and early capital are coalescing. An inflection is possible, but the writings supply no matched indicators proving the gap closed. [researcher inference · strong] 2023 critique · 2025 report
5. Resilience narrative versus financing/governance failure. The MyGlamm essay treats near-failure and a pivot as evidence of founder resilience; later reporting records Good Glamm's lender-driven breakup. Resilience can be real without assuring durable integration, solvency, or governance. [researcher inference · strong] MyGlamm essay · Economic Times timeline
6. Hands-on differentiation versus capacity and attribution. Erasmic promised wide operating support and listed favorable interventions, but gave no partner-hour budget, adverse cohort, handoff metrics, or individual attribution. The promise becomes decision-grade only when owner, deliverable, founder consent, opportunity cost, and counterfactual are recorded. [researcher inference · strong] Positioning index · Positioning source note · Track Record index · Track Record source note

Decision playbook and modern translation

Historical operating rules supported by the writings

Modern VC translation (as of 2026-08-19)

These translations are analyst-created unless explicitly labeled investor-stated. Joint authorship remains stated in the adjacent text. They are tests to run, not claims that Prakash used them historically or that they generated superior returns.

Best writings and unresolved record

Read these five first

1. BlueStone IPO reflection (2025). The best Prakash-specific investment writing: one named company, local category thesis, founder screen, product/trust mechanisms, full-stack choice, capital-cycle endurance, and a filing-controllable public milestone. Its retrospective selection bias is explicit. [researcher inference · strong] Essay
2. Erasmic Funding Philosophy (captured 2008). The closest public period institutional screen: customer evidence, differentiated and capital-efficient service, manager-founder fit, team-specific value add, and sustaining revenue. Read as contemporaneous team doctrine with unknown original publication/revision history, not personal authorship or efficacy evidence. [researcher inference · strong] Common Crawl index record · source note
3. Captain Fresh investment essay (2021). The strongest near-decision, post-decision public rationale: explicit fragmentation, perishability, market-learning, founder, and technology/logistics mechanism at Series A. Market figures and team attribution remain bounded. [researcher inference · strong] Essay
4. The Golden Age of Indian Startups, VCs (2020). The broadest account of how the original Erasmic/Accel team interpreted India's adoption, consumption, B2B, and SaaS opportunity. Read as joint firm history, not Prakash-only attribution or audited performance. [researcher inference · strong] Essay
5. India’s Advanced Manufacturing Leap (2025). The latest located and clearest taxonomy for translating macro enthusiasm into product, equipment, process, IP, and performance questions. Promotional forecasting and coauthorship require skepticism. [researcher inference · strong] Report

Unresolved record and retrievals that could change the synthesis

• Obtain Unboxing Bengaluru through purchase or legitimate library borrowing; separate Prakash's contribution from Malini Goyal's and do not infer content from publicity.
• Retrieve a stable full Decoding Omnichannel file and the underlying forecast/case data; add failed or reversed channel expansions.
• Preserve independent copies or stable archive routes for the nine recovered Erasmic pages; Common Crawl currently supplies the full bodies, but Wayback CDX remained unavailable and the original site is dead. Continue identity-verified social discovery beyond the two processed LinkedIn posts without claiming a complete feed.
• Search private or archival investment memoranda, annual LP letters, IC records, board materials, pass/write-off ledgers, and reserve decisions. The public writings do not establish pricing, ownership, portfolio construction, governance interventions, exits, or personal/fund returns.
• Test the 2023–25 AI/manufacturing forecasts against dated neutral statistics and a complete outcome denominator rather than selected examples.
• Separate sentence-level or section-level authorship in every joint report and investment announcement. Public endorsement is supported; individual composition is not.
• Reconcile the Good Glamm failure and any other adverse outcomes with the original founder, channel, integration, financing, and governance claims before promoting them into mental models.

Coverage and evidence base

The TALKS corpus reconciles 74 deduplicated priority appearance families: 52 ★★★ and 22 ★★, with one notes/t-*.md file for every included family. 38 families expose a full public article, Q&A, publisher transcript, platform transcript, or caption track spanning essentially the complete runtime; 36 remain partial because only publisher highlights, metadata, an indexed body, an incomplete or unusable caption track, or an event summary was recoverable. Contemporaneous record · strong for the local inventory; researcher inference · strong for family boundaries and materiality decisions. Source map

The denominator is deliberately broader than the 20 routes mapped by SCOUT. The TALKS sweep added edited interviews, full platform transcripts, long-form YouTube appearances, conference panels, direct event reports, and first-party SeedToScale episodes across 2008–2026. Duplicate uploads are one family: the two PTI YouTube IDs carry the same 27:15 program, the two Sahyadri videos form one two-part interview, and short clips from longer FCTC or summit recordings are not new appearances. A page, clip, or conference agenda clears the denominator only where Prakash contributes materially; host-only Biopeak introductions, moderator-only sessions, trailers, shorts, reposts, and metadata-only attendance are retained in the exclusion ledger rather than used to inflate coverage. Researcher inference · strong. PTI family note · Sahyadri family note · Source map

The transcript hierarchy was exhausted source by source. Official article/Q&A bodies and platform transcripts were preferred; creator captions and official audio followed; public timestamped auto-caption mirrors were used only when YouTube returned its sign-in/bot challenge; publisher highlights or event recaps became explicitly partial fallbacks. Raw transcripts, captions, audio, and video are not stored in the repository. Machine wording is never promoted above its provenance, speaker separation is preserved in joint sessions, and a missing transcript produces no invented quotation. Contemporaneous record · strong for the retrieval ledger; researcher inference · strong for the evidentiary boundary. Indian Silicon Valley note · COVID publisher-highlights note · CMU partial note

This is a rich but interested record. Prakash speaks variously as an Accel partner, former operator, Erasmic cofounder, ACT Grants participant, civic and policy adviser, philanthropist, personal or vehicle-unclear investor, and longevity initiative patron. Repetition establishes consistency, not causal efficacy or returns. A speaker's Accel title does not turn ACT grants, government work, civic advocacy, or the Biopeak financing into Accel activity; co-panelists' language is not assigned to him. Researcher inference · strong. ACT note · Biopeak investment note · FCTC part 2 note

Source-by-source guide

Two inspected appearances are explicit exclusions. Self Care — the Foundation for Success (uploaded 2025-09-22; 1:01:40) exposed only title/date/runtime metadata and no evidence sufficient to establish Prakash's material contribution. Biopeak Live episode 11 with Matt Kaeberlein (uploaded 2026-05-25; 57:46) contains Prakash's host questions and transitions, while Kaeberlein supplies the scientific, clinical, and entrepreneurial claims; no Prakash investment or career-decision evidence appears. Shorts, trailers, reposts, and duplicate uploads are likewise excluded or consolidated. Researcher inference · strong for the materiality decisions; contemporaneous record · strong for the inspected metadata and caption record. Source map

• [contemporaneous record · moderate] 2008-11-14 — We Are Yet To See Slowdown In Business Plans We Receive. [investor-stated · strong] Prakash said the successor fund would retain Erasmic's seed and early-stage emphasis while using Accel's information and relationship networks to help portfolio companies recruit, sell, and form strategic partnerships. [researcher inference · strong] The interview is contemporaneous with the crisis but success-selected and self-reported: “more than 20” companies is not reconciled to a legal vehicle or full denominator, and the breakeven/profitability claims omit weak, failed, or consolidated positions. Local note
• [contemporaneous record · moderate] 2012-03-29 — We may be 5–7 years from creating ten $1-billion internet firms in India. [investor-stated · moderate] Prakash’s 2012 internet thesis joins increasing broadband, consumer spending, and access: online catalogs could overcome the constrained assortment of physical retail in tier-II and tier-III cities. [contemporaneous record · moderate] Search indexing exposed the full interview body, but the direct publisher page returned a cache miss and curl yielded no readable body in this run. Local note
• [contemporaneous record · moderate] 2012-04-30 — "Early Stage Focus Will Remain, But More Money At Our Disposal". [investor-stated · moderate] Prakash presented the Erasmic–Accel combination as an institutional-capability decision: Accel's domain expertise, global founder appeal, and cross-hub relationships would supplement a locally led India team. [contemporaneous record · strong] The rendered 2012 date conflicts with the interview's 2008 future tense and the independently reported 2008 Fund II chronology. Local note
• [contemporaneous record · moderate] 2012-06-13 — “No more e-commerce” — reported 2012 seed-funding pause. [retrospective witness · provisional] A contemporary Venture Intelligence post reproduces Prakash saying Accel would stop seed-funding new e-commerce trading platforms because “me too” entry had crowded the space, while existing companies might scale, consolidate, or attract later-stage capital. [contemporaneous record · provisional] Only a secondary excerpt is available. Local note
• [contemporaneous record · moderate] 2015-09-14 — People are focusing on mobile-first or mobile-only approaches. [investor-stated · strong] Prakash says Accel did not foresee Indian e-commerce reaching its eventual scale when it first invested: consumer-adoption data were absent, connectivity was weak, and prior attempts had failed. [researcher inference · strong] The headline's mobile-first thesis is only one dated opportunity area, whereas the richest evidence concerns uncertainty, founder observation, and financing power. Local note
• [contemporaneous record · moderate] 2015-12-01 — The kick-start fund. [investor-stated · strong] Prakash described Accel's method as theme-based: select what it regards as the best team within a sector and “almost never” invest in conflicting companies. [contemporaneous record · strong] The “almost never” conflict claim coexists with three reported portfolio combinations. Local note
• [contemporaneous record · moderate] 2016-02-19 — The era of lead gen is game over; have full control of customer engagement. [investor-stated · strong] Prakash expected entrepreneurs starting in 2016 to face more circumspect investors and a different fundraising reality, with a possible 12–24-month lull. [researcher inference · moderate] Saying there was no bubble while calling the period a hyper-valuation cycle and predicting a funding lull preserves a real tension: Prakash expected repricing without systemic collapse. Local note
• [contemporaneous record · moderate] Circa 2017 — A Parallel Universe — BECO Capital panel with Vani Kola and Prashanth Prakash. [investor-stated · moderate] Prakash says his first India venture was an early internet company and that he and his business partner both exited companies in the same quarter. [researcher inference · strong] Fund origin, Google commitment, portfolio scale, and LP concentration are retrospective self-reports without governing documents or a named fund vintage. Local note
• [contemporaneous record · moderate] 2018-01-10 — How to Start a Start-up and Digitalization of the Economy — Sahyadri Conclave 2018. [contemporaneous record · strong] Sahyadri's event report says Prakash's session explained how digitalization transformed industry, defined transformation and subsets of digitalization, and discussed data analytics, personalization, and network effects. [contemporaneous record · strong] The videos are publicly indexed, but yt-dlp hit YouTube bot confirmation, transcript clients were blocked, timed-text returned no usable body, and no legitimate mirror was found. Local note
• [contemporaneous record · moderate] 2018-05-11 — Markets Opportunities — Evaluate Wise & Execute Nice. [investor-stated · moderate] Prakash frames the first Indian consumer-internet opportunity as making access predictable: a venture must own the operational work required to deliver a consistently good customer experience, not merely add a digital interface. [contemporaneous record · strong] SeedToScale exposes the full official audio and a “Download Transcript” control, but the control resolves only to the page anchor. Local note
• [contemporaneous record · moderate] 2018-07-13 — The Accel effect. [investor-stated · strong] His selection screen combined fund-scale relevance, an asset-light and technology-enabled model, timing, and a long period spent forming a view of the market and company. [researcher inference · strong] The article's celebratory title and closing virtuous-cycle forecast coexist with its own harder evidence: notional marks, stock consideration, estimated rather than audited multiples, material company losses, and a persistent exit bottleneck. Local note
• [contemporaneous record · moderate] 2018-07-14 — Evening Keynote at the CMU India Conclave. [contemporaneous record · strong] Carnegie Mellon University's official channel titles the 56:03 item an “Evening Keynote by Prashanth Prakash” at the CMU India Conclave in Bangalore on July 14, 2018. [contemporaneous record · moderate] CMU's item calls Prakash's session an evening keynote, while Deccan Herald calls Devi Shetty “the keynote speaker” and lists Prakash among other speakers. Local note
• [contemporaneous record · moderate] 2019-06-28 — Entrepreneur India Conclave 2018 — sector and selection remarks. [investor-stated · strong] Prakash reduces early-stage selection to three variables: latent/developing market opportunity, entrepreneurial-team quality, and early evidence of execution. [contemporaneous record · moderate] The source is a multi-investor retrospective with edited excerpts and apparent transcription errors, not an official full transcript or recording. Local note
• [contemporaneous record · moderate] 2020-01-15 — Accel Partners: “We don't see us as investors, but co-creators” — YourStory. [contemporaneous record · strong] YourStory frames Accel India's operating ethos as co-creation: work from the ground up with founders, add value across stages, and build trust rather than function only as a capital provider. [researcher inference · strong] “Co-creator” can describe genuine operating help or retrospectively enlarge investor credit. Local note
• [contemporaneous record · moderate] 2020-01-16 — Accel India founders on investing through instinct and intuition. [investor-stated · strong] Prakash's only substantive published excerpt says instinct and intuition must be used “in a guarded manner.” He warns that even deep familiarity with a space can create conditioning that becomes counterproductive, and connects the freedom to hold uncertainty with seed investing. [researcher inference · strong] The article headline and standfirst jointly attribute instinct and intuition to both founders, but the body gives different claims to each speaker. Local note
• [contemporaneous record · moderate] 2020-04-17 — The State of the VC Ecosystem amidst the COVID-19 Crisis. [contemporaneous record · strong] The publisher structures the conversation from historical comparison (dot-com and 2008) through ecosystem response, portfolio risk, founder survival, pre-product-market-fit choices, a 24-month outlook, reinvention, sector change, and anxiety. [contemporaneous record · strong] The official audio enclosure and YouTube video are public, but yt-dlp returned YouTube’s “Sign in to confirm you’re not a bot” barrier, the public timed-text endpoint returned no caption body, and publisher/Apple/Spotify/RSS routes exposed show notes rather than a transcript. Local note
• [contemporaneous record · moderate] 2020-08-14 — ACT Grants helped create disruptive, scalable solutions to fight Covid-19. [investor-stated · moderate] Prakash says ACT emerged after more than 1,000 startup participants gathered in a Telegram group and 15–20 venture, private-equity, and investment-banking groups created a structured grant mechanism. [researcher inference · strong] Venture vocabulary—IC, scalable technology, teams, and market readiness—does not turn grants into investments. Local note
• [contemporaneous record · moderate] 2020-11-22 — Investments From US, Middle East Set To Make Up For Chinese Capital. [investor-stated · moderate] Prakash described the first pandemic months as an observation period and estimated a six-month investment-market lag. [researcher inference · moderate] Vedantu and Ninjacart are two favorable examples within an undisclosed portfolio. Local note
• [contemporaneous record · moderate] 2020-11-26 — Investments will follow good entrepreneurs in agritech — TechCircle interview. [investor-stated · moderate] Prakash attributes agritech's improving opportunity set to smartphone adoption and asset-light technology companies tackling agricultural bottlenecks. [researcher inference · strong] Execution quality is asserted as the capital magnet, but the appearance record supplies no operational KPI, founder-assessment process, or evidence separating execution from market timing and follow-on capital availability. Local note
• [contemporaneous record · moderate] 2020-12-11 — Future of Agritech: The evolution of Agritech in India. [contemporaneous record · strong] The panel’s public agenda divides the market into landscape, funding tailwinds, farm bills, COVID effects, farm-to-consumer, precision agronomy, and founder advice. [contemporaneous record · strong] The official video and audio are accessible, but yt-dlp hit YouTube’s bot-confirmation barrier, the public timed-text endpoint returned no captions, publisher/Apple/Spotify searches exposed no transcript, and the page’s transcript control is a non-file anchor. Local note
• [contemporaneous record · moderate] 2021-07-01 — A VC's Tips to Survive & Thrive in the Digital Age — Spotlight with Sandhya. [investor-stated · strong] Prakash treats the pandemic as an accelerator of existing digital marketplaces, payments, and fintech: changing customer behavior pulled some businesses roughly two years ahead of prior projections. [contemporaneous record · strong] Public YouTube transcript and media routes were blocked by bot confirmation, but the publisher's RSS feed exposed the official audio enclosure. Local note
• [contemporaneous record · moderate] 2021-08-06 — Future of Beauty and Personal Care: Tracking the evolution of the category in India. [contemporaneous record · strong] The publisher isolates Prakash’s category framing at 06:30–10:51, investor perspective at 20:50–24:00, and offline-survival point at 30:35–31:45. [contemporaneous record · strong] The official video/audio and chapter index are accessible, but yt-dlp met YouTube’s bot-confirmation barrier, the public timed-text endpoint returned no caption body, and publisher/Apple/Spotify/RSS searches produced no transcript. Local note
• [contemporaneous record · moderate] 2021-11-16 — Pioneering Indian Venture Capital — Indian Silicon Valley. [investor-stated · strong] Prakash defines venture as a five-to-ten-year foresight exercise, but constrains the investor's role: a prepared mind should identify an emerging space and know what to look for, while the entrepreneur remains better placed to decide exactly what to build. [researcher inference · strong] The interview's “prepared mind” narrative can make successful category sequences look planned after the fact. Local note
• [contemporaneous record · moderate] 2021-11-17 — Will look at global crypto opportunities; there are regulatory challenges in India. [investor-stated · moderate] At the 2021 funding peak, Prakash saw internet IPOs as a catalytic proof of Indian exit capacity that could attract more capital and support experiments in longer-cycle space and climate technology. [contemporaneous record · strong] The full edited text is available, but no audio or verbatim transcript is linked; editorial omissions and paraphrase outside answer blocks cannot be reconstructed. Local note
• [contemporaneous record · moderate] 2022-10-17 — India's Startup Story at Khazanah Megatrends Forum 2022. [investor-stated · moderate] Prakash divided India's technology-company formation into IT-services foundations, global product and engineering centers, and a startup/digitization phase. [researcher inference · moderate] The presentation celebrates 100 unicorns while also saying profitable, sustainable startup construction remained unsolved. Local note
• [contemporaneous record · moderate] 2022-11-16 — Indian startups need to deliver good profits in three to five years. [investor-stated · strong] Prakash distinguishes headline valuations from durable venture outcomes: good profits and public exits are necessary to return capital to investors and sustain later inflows. [contemporaneous record · moderate] The article preserves selected remarks but not a complete event recording or transcript. Local note
• [contemporaneous record · moderate] 2022-11-17 — Epicentre Of Climate-Change Tech Will Move Into India. [investor-stated · moderate] Prakash said ACT grew from a COVID-response coalition into a multi-cause platform after participants saw startup-built solutions as applicable to broader social problems. [researcher inference · strong] A voluntary pledge plus resources can encourage action, but neither signatories nor tools prove emissions reduction. Local note
• [contemporaneous record · moderate] 2022-11-29 — How Infra.Market offers concrete solutions through its construction marketplace. [retrospective witness · moderate] Founder Souvik Sengupta frames the opportunity as a two-sided fragmentation problem: construction companies repeatedly rebuild local procurement networks, while small producers lack national distribution. [researcher inference · strong] “Asset-light” and “quasi ownership” pull in opposite directions: low fixed investment depends on monitoring, contracts, counterparty cooperation, and enforcement that can reintroduce operational and balance-sheet risk. Local note
• [contemporaneous record · moderate] 2022-12-15 — Borderless Ambitions — Globalising the India Stack. [investor-stated · moderate] Prakash used Juspay to explain a monetization layer around nominally free UPI transactions: banks and other providers pay for enabling services, while payment apps such as PhonePe can monetize sticky customers through insurance and other products. [researcher inference · moderate] “UPI itself is free” requires a scope boundary: end-user transfers can be free while banks, apps, infrastructure providers, or merchants still bear and monetize surrounding costs. Local note
• [contemporaneous record · moderate] 2022-12-15 — The Start-up Pledge for Climate Action. [investor-stated · moderate] Prakash separated his Accel role from ACT, describing ACT as a coalition of venture funds, technology entrepreneurs, investment bankers, and social-sector leaders formed during COVID and then extended to climate, education, and healthcare. [contemporaneous record · strong] ACT's June 2023 write-up says implementation would begin with 6–10 startups, and a later ACT newsletter describes a 10-startup pilot with StepChange. Local note
• [contemporaneous record · moderate] 2023-01-07 — Indian Startups: Will 2023 Be Better Than 2022? — CNBC-TV18 Young Turks. [contemporaneous record · strong] CNBC-TV18 identifies the appearance as a January 7, 2023 Young Turks discussion in which Sequoia India managing director Mohit Bhatnagar and Accel partner Prashanth Prakash shared their outlook for Indian startups in 2023. [contemporaneous record · strong] YouTube displays a transcript control, but the public transcript client, timed-text endpoint, and yt-dlp media route were blocked or empty; web search found no transcript mirror or speaker-attributed recap of this exact interview. Local note
• [contemporaneous record · moderate] 2023-04-03 — Coffee with Ashwath — startups, collective philanthropy, and co-building. [investor-stated · strong] Prakash describes ACT's pandemic model as a coalition of investors/donors, NGOs, and startup founders with technology and operating ideas. [researcher inference · strong] Startup speed and government scale can complement each other, but the analogy may obscure procurement, evidence quality, public accountability, maintenance, and community consent. Local note
• [contemporaneous record · moderate] 2023-06-02 — There is no real ecosystem for deep tech in India. [investor-stated · strong] Prakash’s 2023 deep-tech diagnosis is ecosystem-wide: modest corporate and government R&D, an early financing cliff after $1–3 million, weak university translation incentives, missing industry links, and poor acquisition demand for intellectual property. [contemporaneous record · strong] The article contains selected direct quotations, not an official transcript. Local note
• [contemporaneous record · moderate] 2023-08-06 — Prashanth Prakash in Conversation at Venture Catalysts' Bangalore Office Launch. [investor-stated · moderate] Sharma's contemporaneous post attributes to Prakash a retrospective arc from entering Indian venture in 2008 without a playbook to a major Flipkart exit in 2023. [researcher inference · strong] “Best entrepreneurs globally” is not a testable conclusion without a comparison set and metric; the evidence offered is a successful case, not a global founder-quality distribution. Local note
• [contemporaneous record · moderate] 2023-09-01 — Ep #9: WTF Is Venture Capital?. [contemporaneous record · strong] The official Nikhil Kamath title and Blume's participant page identify a venture-capital roundtable with Nikhil Kamath, Nithin Kamath, Rajan Anandan, Prashanth Prakash, and Karthik Reddy. [contemporaneous record · strong] Public editions differ by one day in displayed date and by seconds in displayed duration: YouTube shows a September 1 upload and 2:29:40 runtime, while Apple and Spotify date the episode September 2 and Spotify shows 2:29:39. Local note
• [contemporaneous record · moderate] 2023-11-21/22 event window; exact session/upload date unavailable — TIA Summit 2023 — Unboxing Bengaluru and China-plus-N manufacturing. [investor-stated · moderate] Prakash treats Bengaluru as evidence that innovation has dispersed beyond Silicon Valley. [researcher inference · strong] The 60–80% Bengaluru claim, “first” claims, fund scale, company count, and unicorn count are promotional self-reports with inconsistent possible definitions. Local note
• [contemporaneous record · moderate] 2023-11-28 — Unboxing Bengaluru: The City of New Beginnings — India Global Forum. [investor-stated · strong] Prakash frames Bengaluru as a bounded proxy for understanding India's last three to four decades of transformation. [researcher inference · strong] The proxy claim risks selection bias: Bengaluru is unusually wealthy, technical, migratory, and English-accessible, so its consumers and institutions may not represent India outside comparable urban clusters. Local note
• [contemporaneous record · moderate] 2023-12-21 — Karya AI and the Potential of AI to Uplift Bharat — Global AI Conclave. [investor-stated · strong] Prakash argues that AI for lower-income and lower-literacy users needs simpler interfaces, especially voice interaction in Indian languages, because conventional search can be intimidating. [researcher inference · moderate] “Global first, adapt for Bharat” can conflict with designing around low literacy, voice, language, connectivity, and willingness-to-pay from inception. Local note
• [contemporaneous record · moderate] 2023-12-28 — Invest in new ideas, not flavour of the season. [investor-stated · strong] Prakash says Accel maintains “prepared minds,” reassessing spaces every four or five years and studying themes four or five years before expected impact. [contemporaneous record · strong] The article is full text but explicitly edited; omitted questions, answer cuts, and recording context are unavailable. Local note
• [contemporaneous record · moderate] 2024-03-25 — Accel aims to invest in 15–25 Indian startups in 2024. [investor-stated · strong] Prakash said Accel still had substantial dry powder, expected 15–25 Indian investments in 2024, and generally launched a fund every three to four years after deploying the current vehicle. [researcher inference · strong] A longer process can mean better diligence, slower decision-making, or both. Local note
• [contemporaneous record · moderate] 2024-05-10 — Prashanth Prakash: Master mentor. [investor-stated · strong] Prakash describes venture work as forming forward-looking views, locating entrepreneurs who can embody them, and serving as a guide whose involvement can range from operational detail to strategy when invited. [contemporaneous record · strong] The article is fully readable, but it is a reporter-shaped profile rather than a verbatim transcript; only text inside quotation marks is treated as direct voice. Local note
• [contemporaneous record · moderate] 2024-05-10 — There are not enough consumer brands; we have only seen the tip of the iceberg. [investor-stated · strong] Prakash says consumer brands should scale through online and offline platforms “wisely and profitably”; the tipping point is brand formation, not permission to subsidize any omnichannel expansion. [contemporaneous record · strong] The edited text is complete and speaker-labelled but no recording or verbatim transcript is linked. Local note
• [contemporaneous record · moderate] 2024-06-10 — A Toolkit for Decoding Omnichannel — Accel India launch conversation. [investor-stated · strong] Prakash distinguishes the launch document from a conventional market report: after a landscape of “trailblazers,” it is meant to help brands decide readiness, choose an offline format, pilot, select metrics, and graduate toward national or multi-format expansion. [researcher inference · strong] A toolkit can discipline expansion but may create false precision across categories with different margins, replenishment, discovery, return rates, and working-capital cycles. Local note
• [contemporaneous record · moderate] 2024-08-14 — Should founders move from Bengaluru to Mumbai to build? — Prashanth Prakash and Sanjay Nayar. [investor-stated · strong] Prakash says the first startup IPO cohort disappointed, but founders and investors quickly learned the public-market discipline: constrain promises, deliver consistently, and combine growth with profits. [researcher inference · strong] Prakash's public-market optimism is conditional on a scarce qualifying asset pool; that scarcity can support valuations but can also indicate that venture-backed companies have not yet met his own revenue-and-profit test. Local note
• [contemporaneous record · moderate] 2024-10-11 — TIA Summit 2024 — domestic capital, frugality, and Global South platforms. [investor-stated · provisional] Prakash compares India's current startup phase with China's 2008–18 decade and argues that domestic capital—not foreign capital alone—is the step-change because it is persistent and locally committed. [researcher inference · strong] Monthly public-market inflows are not the same as domestic venture commitments. Local note
• [contemporaneous record · moderate] 2024-11-24 — Robotics, Precision Manufacturing Hot Bets For Investors. [investor-stated · moderate] Prakash identified robotics, precision manufacturing, and combinations of software with physical production as near-term founder and venture opportunities. [researcher inference · strong] The “seven times capacity” comparison is not decision-ready because neither demand nor capacity is defined. Local note
• [contemporaneous record · moderate] 2024-12-10 — Is AI really going to force Indians out of jobs? — FCTC episode 6, part 1. [investor-stated · moderate] The episode is principally a formative-biography and civic-operating interview, not an answer to its AI-and-jobs headline. [contemporaneous record · strong] Despite the headline, the complete mirrored captions contain no sustained analysis of Indian job displacement. Local note
• [contemporaneous record · moderate] 2024-12-17 — How this VC straddles startups, politics and development — FCTC episode 6, part 2. [investor-stated · strong] Prakash opens the institutional history by rejecting a founder-hero account: he says Erasmic required three complementary co-founders and that Accel India would not have existed without that combination. [researcher inference · strong] The host introduces Prakash as foundational to Flipkart, while Prakash's own detailed account shifts decisive proximity to Goyal, board involvement to Mitra, regulation to Balachandran, and the investment to a team. Local note
• [contemporaneous record · moderate] 2024; exact date unavailable — Demo Day 2024 keynote — Ramaiah Evolute. [investor-stated · strong] Prakash rejects unicorn size as the only valid outcome. [researcher inference · strong] The IPO forecast is precise enough to test but underspecified: “startup,” listing venue, completion versus filing, India origin, and calendar cutoff are undefined. Local note
• [contemporaneous record · moderate] 2025-02-10 — “We Expect 70–100 IPOs This Year,” Says Accel's Prashanth Prakash. [investor-stated · provisional] Prakash characterized India as an unusually complete emerging-market startup ecosystem because startups span sectors and reinforce capital and entrepreneurship. [researcher inference · strong] The article's introductory “funding recovery” frame is compatible with Prakash's stage-specific caution only if recovery is uneven: he simultaneously describes smaller early rounds and a two-to-three-year Series B/early-growth crunch. Local note
• [contemporaneous record · moderate] 2025-04-05 — Are Indian Startups Gaining from the New US Tariffs? — Startup Mahakumbh 2025. [contemporaneous record · strong] TICE TV identifies Prakash as the interview subject and says he gives a “deep dive” into Startup Mahakumbh 2025's planning, purpose, first-day turnout, ecosystem implications, and next-year vision. [contemporaneous record · strong] The YouTube page advertised a transcript, but the transcript body resolved empty; the official caption client was request-blocked, and the media route required sign-in to confirm a viewer was not a bot. Local note
• [contemporaneous record · moderate] 2025-04-08 — Venture debt across the startup capital continuum. [investor-stated · moderate] Prakash says Indian venture debt evolved from bridge/runway extension into working-capital, supply-chain, early bank-relationship, global-expansion, asset-based, and pre-IPO uses. [researcher inference · strong] Debt can reduce equity dilution but adds repayment, covenant, security, and refinancing risk. Local note
• [contemporaneous record · moderate] 2025-04-14 — India Can Become a Global Hub for IP-Led Innovation. [investor-stated · strong] Prakash said India could become a hub for IP-led innovation and high-value manufacturing in defence, electric vehicles, aerospace, and energy. [contemporaneous record · strong] The published PTI text and Prakash's own post specify a $650 million fund, while the third-party automatic captions appear to render the spoken figure as $700 million near 20:30. Local note
• [contemporaneous record · moderate] 2025-05-20 — Karnataka fintech, talent, and matched startup funding — World Fintech Summit 2025. [investor-stated · strong] His first institutional proposal is to use the fintech center of excellence for specialized skilling in AI and cybersecurity, so the existing cluster can preserve its talent advantage as fintech becomes more technically demanding. [researcher inference · strong] The 50–60% and 60%-plus shares lack company lists, geographic rules, revenue definitions, or de-duplication across payments, broking, and insurance; they are positioning claims, not established market measurements. Local note
• [contemporaneous record · moderate] 2025-06-06 — Over 250 Indian AI startups and a voice-first opportunity. [investor-stated · moderate] Prakash's stated adoption thesis is that voice is the natural interface for the next billion users and that multilingual Indian-language capability can broaden access. [researcher inference · strong] A large startup count measures formation, not quality, survival, revenue, ownership, or return. Local note
• [contemporaneous record · moderate] 2025-06-10 — Accel's AI capital allocation and the IndiaAI Mission. [investor-stated · strong] Prakash says Accel remains primarily an early-stage investor—seed and Series A—with selective growth investments, and that this stage posture would continue for its AI-focused capital. [researcher inference · strong] Forty companies in 18 months signals portfolio breadth but says nothing about quality, ownership, capital concentration, survival, or returns. Local note
• [contemporaneous record · moderate] 2025-07-28 — India can be a global lab for scalable longevity solutions. [investor-stated · moderate] Prakash frames the Biopeak decision around the convergence of diagnostics, AI, and longevity science, with a precision-health system adapted to Indian biology, environment, and lifestyle. [researcher inference · strong] The interview's scientific and accessibility language is investment advocacy from an investor in the company. Local note
• [contemporaneous record · moderate] 2025-08-05 — More capital can compress advanced-manufacturing gestation periods. [investor-stated · moderate] Prakash gives four reasons for Accel's advanced-manufacturing focus: China+1 supply-chain movement, a policy transition from production incentives toward design incentives, more engineering-first founder inbound, and a large market forecast. [researcher inference · strong] More capital can fund parallel development and capacity, but it cannot by itself compress physical validation, certification, procurement, or regulatory time. Local note
• [contemporaneous record · moderate] 2025-08-05 — Accel rejigs India playbook after previous misses. [investor-stated · strong] Prakash openly says Indian climate technology had not come close to its potential; Mint groups sustainability, agritech, and electric vehicles among Accel theses that underdelivered. [contemporaneous record · strong] The readable article is a two-minute reported interview, not a complete transcript. Local note
• [contemporaneous record · moderate] 2025-08-06 — India will see advanced-manufacturing unicorns within three to four years. [investor-stated · moderate] Prakash said venture attitudes toward defence, robotics, semiconductors, aerospace, and other asset-heavy businesses had shifted in the preceding six to nine months. [researcher inference · moderate] Prakash's 2023 claim that India lacked a real deep-tech ecosystem is not necessarily inconsistent with a 2025 inflection claim, but the later source must identify what changed and measure it rather than rely on sentiment. Local note
• [contemporaneous record · moderate] 2025-09-11 — Pitching Jewelry for the Digital Age: Inside BlueStone’s Original Deck. [contemporaneous record · strong] The 2011 deck’s ex-ante thesis combined high ticket size and margins, a shift toward branded/lightweight/fashion jewelry, a nascent online channel, and an expected low-inventory model. [contemporaneous record · strong] The publisher calls the source a 33-page idea-stage deck, while the downloadable PDF has 25 PDF pages: slides 1–24 followed by slide 33. Local note
• [contemporaneous record · moderate] 2025-09-22 — H-1B fee increase as a possible positive for Indian startups. [investor-stated · moderate] Prakash viewed the announced H-1B fee increase as a potential positive for Indian venture and startups because some U.S.-experienced product and AI talent might return to India. [researcher inference · strong] A policy that may benefit India's talent supply can still harm individual workers, students, U.S. startups, and cross-border knowledge exchange. Local note
• [contemporaneous record · moderate] 2025-09-25 — Medicine 4.0: Innovations for the Future of Indian Healthcare. [contemporaneous record · strong] Longevity India identifies Prakash and Nithin Kamath as the two fireside participants and says both unpack Medicine 3.0's infrastructure and personalization shortcomings. [contemporaneous record · strong] The organizer's May 30 recap predates the September 25 upload and separately matches its main joint summary, but both are Longevity India publications; this is same-organizer corroboration, not source independence. Local note
• [contemporaneous record · moderate] 2025-10-06 — Unboxing Bengaluru: Memory, Metropolis, and the Making of a New City — BBLF2025. [investor-stated · strong] Prakash declares the urban-convenience startup era mature and argues Bengaluru's next reinvention should be “industrial AI”: full-stack companies combining software, AI, hardware, and scaled manufacturing. [researcher inference · strong] The Flipkart origin story compresses many causal contributors into Amazon location, founder migration, and Accel backing; the panel provides no contemporaneous memo, cap-table evidence, or alternative witness account. Local note
• [contemporaneous record · moderate] 2026-01-04 — India's evolving ecosystem — SuperReturn Asia fireside. [contemporaneous record · moderate] Interviewer Arnaud Bonzom and Black Mangroves summarize Prakash as arguing that India's ecosystem is approaching “completeness,” with domestic B2C IPOs adding to U.S.-listed SaaS precedents. [researcher inference · strong] The no-foundation-model summary sits in tension with Prakash's June 2025 openness to monetizable sovereign and India-specific models. Local note
• [contemporaneous record · moderate] 2026-01-07 — Data Science & AI for Longevity. [contemporaneous record · strong] Longevity India says Prashanth Prakash and AI/ML specialist Ram Prakash H jointly explore data science and AI in longevity and preventive care. [researcher inference · strong] The publisher's systems-safety summary is directionally cautious, but it is not an independent technical or clinical evaluation. Local note
• [contemporaneous record · moderate] 2026-01-10 — Shift towards profitability, not a funding winter. [investor-stated · strong] Prakash distinguishes a stage and sector reallocation from a generalized funding winter: he sees adequate $3–8 million pre-Series A and Series A capital in AI, deep tech, and some consumer segments, but a shortage in $20–50 million Series B/C rounds. [contemporaneous record · strong] The body is an edited Q&A preserved in page metadata, not a recording or verbatim transcript. Local note
• [contemporaneous record · moderate] 2026-03-04 — From 500 to 2 Lakh Startups: How India Is Becoming a Global Tech Superpower. [investor-stated · moderate] Prakash rejected a lone-investor version of Accel's successes, describing the named companies as firm and team investments. [contemporaneous record · strong] The title's “500 to 2 Lakh Startups” and several ecosystem totals are introduced by the interviewer. Local note
• [contemporaneous record · moderate] 2026-04-14 — The Journey of Generosity at the 2026 Indiaspora Forum. [contemporaneous record · strong] Indiaspora explicitly lists Prakash, Arun Jain, and Latha Rajnikanth as panel speakers while separately identifying Atul Satija as session chair. [contemporaneous record · strong] The video page exposed a transcript control but no transcript body; the official caption client was request-blocked and media retrieval was bot-blocked. Local note
• [contemporaneous record · moderate] 2026-05-14 — Value, Valuation & the Void in Between. [investor-stated · moderate] Prakash treated direct-to-consumer as an entry point for building a brand around digital-first, trend-setting customers rather than a permanent channel definition. [researcher inference · strong] The session title promises value and valuation, but Prakash offers outcome heuristics rather than a valuation method: no discount rate, ownership, dilution, margin, growth, comparable, return hurdle, or probability model is disclosed. Local note
• [contemporaneous record · moderate] 2026-05-19 — Becoming an Inclusive $30 Trillion Economy by 2047. [investor-stated · moderate] After an explicit handoff, Prakash said investors should ask where trillions of dollars of economic activity can come from and what gives India a “right to win.” His proposed foundation was dual-use technology and advanced manufacturing, not defence alone. [contemporaneous record · strong] The $30 trillion target belongs to the session title and publisher framing. Local note
• [contemporaneous record · moderate] 2026-06-02 — We Can Do Much More For Our Country — India Philanthropy Summit 2026 fireside chat. [investor-stated · strong] In the accessible excerpt, Prakash distinguishes emerging philanthropy from passive cheque-writing: he says the next generation wants direct involvement, institution-building, disruption, and scalable solutions. [researcher inference · strong] “Disruption,” “innovation,” and “scale” are directional rhetoric without a named cause, theory of change, beneficiary-governance model, risk budget, or measured outcome in the accessible excerpt. Local note
• [contemporaneous record · moderate] Undated — Guest of the Week — Prashanth Prakash on Startup Karnataka. [contemporaneous record · strong] Doordarshan Chandana presents Prashanth Prakash as the sole “Guest of the Week” and identifies him as chairman of the Vision Group on Startup. [contemporaneous record · strong] The YouTube watch page and transcript client were blocked by bot confirmation; the public timed-text endpoint returned no usable captions; search results exposed only title, runtime, channel, and description. Local note
• [contemporaneous record · moderate] Undated — Grounding Innovation in Spirituality — INKtalks. [investor-stated · moderate] Prakash frames the Nandi-area venue as a deliberate alternative to private consumption: a place for spiritual, creative, and immersive experiences rather than merely a large house. [researcher inference · strong] The geological-age phrase is almost certainly misrecognized by ASR (“million” where the geological context may require another unit). Local note

Recurring decision frameworks

1. Prepared mind, then non-consensus opportunity

• [investor-stated · strong] Prakash repeatedly rejects investing by current fashion. The durable version is not clairvoyance: form a view years ahead, keep enough domain context to recognize a team or market inflection, and remain willing to update when a founder's view is better. His 2015 Flipkart answer is especially valuable because he admits there was no consumer-adoption data and that Accel did not anticipate the eventual scale. Invest in new ideas · mobile-first/Flipkart interview · Master mentor
• [researcher inference · strong] “Prepared mind” and “intuition” are compatible only if intuition is treated as a compressed, challengeable hypothesis—not an exemption from evidence. The public record does not show the complete declined-deal set needed to test whether intuition selected winners or merely explains them in hindsight. Instinct and intuition · Pioneering Indian VC

2. Start with unmet need, reachable customer, and bounded TAM

• [investor-stated · strong] The most portable early-stage screen is a sequence: own predictable customer access and the heavy lifting; identify an unmet need; bound TAM regionally; and test expansion from a tractable wedge into adjacent categories. This is more decision-useful than the broad sector lists that recur in press interviews. Markets & Opportunities
• [investor-stated · moderate] Customer access is not merely lead generation. In 2016 Prakash argued that internet businesses should control customer engagement and retention rather than depend on lead intermediaries or discounts; later omnichannel work turns that into a staged online-to-offline system. Customer engagement · omnichannel interview
• [researcher inference · strong] A large TAM, policy tailwind, or “next billion” interface does not establish a monetizable market. The public talks rarely disclose conversion, price, cohort retention, gross margin, or a rejected-market denominator. AI voice-first remarks · 2012 internet-unicorn forecast

3. Select for founder behavior, learning, and customer obsession

• [investor-stated · strong] The recurring founder signal is behavior under ambiguity: technical capacity combined with willingness to solve logistics, service, packaging, and other unglamorous customer problems. Prakash's Flipkart recollection is strongest as an observation about the Bansals, not as proof of his personal investment lead. Mobile-first/Flipkart interview
• [investor-stated · moderate] The sources separately emphasize mission and domain insight, founder–investor compatibility, and operational resilience. Spotlight's evidence is about operating-team fungibility, channel redundancy, and distributed culture—not founder selection. The record is weak on falsification: no source gives a complete scorecard, rejected founders, founder-reference misses, or rates of post-investment team failure. Master mentor · Indian Silicon Valley · Spotlight

4. Early-stage institutional platform with follow-on capacity, not a lone picker

• [investor-stated · strong] From the 2008 crisis interview through the 2012 Erasmic–Accel discussion, Prakash frames the institutional advantage as local judgment plus global expertise, founder access, larger initial positions, and follow-on capacity. He explicitly declines to disclose combination economics, an important boundary against invented carry or ownership. 2008 business-plans interview · fund-combination interview
• [investor-stated · moderate] By 2024 he describes a three-to-four-year fund cadence, dry-powder-dependent launch timing, and 15–25 planned annual investments. By 2025 the AI clip still says seed and Series A dominate, with selected growth. These are firm-level plans, not a person-level portfolio or cash-flow ledger. Investment pace and diligence · IndiaAI capital allocation
• [researcher inference · strong] More capital has two opposed mechanisms in this corpus: it can preserve ownership and finance winners, but it can also enable overfunding, escalation, and excessive pace. Vehicle terms, reserves, ownership, follow-on loss rates, and vintage results remain missing. Fund-combination interview · manufacturing gestation interview · venture-debt interview

5. Mentor by permission; capacity and governance still bind

• [investor-stated · moderate] Prakash's ideal role is guide and mentor, with involvement ranging from tactical detail to strategy when founders invite it. He resists the “adult supervision” model. Master mentor
• [researcher inference · moderate] His reported 15–18-company active workload and preference for in-person meetings imply a practical capacity constraint; he does not state a board-capacity doctrine. The 2023 Mint interview instead establishes limited unilateral board control. Master mentor · Invest in new ideas
• [investor-stated · moderate] The 2015 joint interview says entrepreneurs cannot be forced because they have choices. The later governance-failure environment leads Accel to nearly double diligence time. The corpus does not show whether the founder-autonomy norm, board intervention, and investor accountability were reconciled in distressed cases. Mobile-first interview · Investment pace and diligence
• [researcher inference · insufficient public record] No complete board calendar, intervention log, founder testimony denominator, recusal policy, or failed-intervention set tests the mentoring claim. Company outcomes cannot be attributed to advice by proximity. Master mentor

6. Capital efficiency, profitability, and financing fit

• [investor-stated · strong] Capital efficiency appears in the 2008 downturn, the 2016 funding lull, the 2022 three-to-five-year profitability expectation, and the 2026 rejection of “funding winter” language. The stable core is not zero burn; it is matching capital, growth, and business model to a credible path to operating durability. 2008 business plans · customer engagement · profitability forecast · 2026 funding interview
• [investor-stated · moderate] Venture debt is presented as a continuum—from runway and working capital to asset finance and pre-IPO uses—rather than a substitute for equity at one stage. The event context is lender-friendly and omits coupon, covenant, warrant, default, and refinancing evidence. Venture-debt continuum
• [researcher inference · strong] Profitability language is vulnerable to outcome selection: slower growth, layoffs, delayed investment, discounted asset sales, or lender pressure can also improve short-term cash metrics. Company filings and full cohorts are required. Kick-start fund · Accel effect

7. Advanced manufacturing: IP, maturity, and global relevance

• [investor-stated · strong] The 2025 manufacturing system is unusually concrete. Accel divides innovation into product, process, and equipment; generally seeks at least a validated TRL-6 prototype plus pilots or early customer traction; and wants global relevance with Indian initial demand, except where defence begins domestically. Manufacturing gestation · previous misses
• [investor-stated · provisional] Prakash expects capital to compress five-to-ten-year gestation into three-to-five years, initial $5–10 million rounds to become $30–50 million rounds within 18–36 months, and an advanced-manufacturing unicorn within three-to-four years. These are dated forecasts, not current facts; certification, procurement, validation, and yield may not compress with financing. Manufacturing gestation · unicorn forecast
• [researcher inference · strong] The 2025 optimism is credible evidence of strategy change because Prakash directly says Indian climate tech had not lived up to potential. Mint, not Prakash, frames sustainability, agritech, and EV as other underdelivering Accel theses. Neither statement shows that the new taxonomy fixes selection, capital intensity, or commercialization. Previous misses

8. AI: application layer, local language, and selective sovereignty

• [investor-stated · moderate] In 2024 FCTC part 2, Prakash treats AI as an enabling component and a requirement for global competitiveness. A 2025 Moneycontrol report attributes compute, data, R&D, and talent constraints jointly to Prakash and a MeitY official, so those constraints cannot be assigned to him alone. FCTC part 2 · AI voice-first
• [investor-stated · moderate] Direct 2025 evidence supports contextual applications, voice and local-language interfaces, and openness to monetizable India-specific models. AI voice-first · IndiaAI capital allocation
• [contemporaneous record · provisional] A 2026 SuperReturn organizer summary reports an application/agent preference and says he doubted India would produce foundation models because of $30–50 billion capital requirements. This may distinguish frontier general models from targeted sovereign models, but the missing recording prevents confident reconciliation. SuperReturn summary
• [researcher inference · strong] Service workflow scale is not automatically proprietary data, reusable IP, distribution, or pricing power. The “40 plus AI companies” and “250 startups” claims have no public deduplicated roster or outcome denominator. IndiaAI capital allocation · AI voice-first

9. Exits and liquidity are ecosystem infrastructure, not paper marks

• [investor-stated · moderate] Prakash repeatedly treats IPOs and M&A as necessary for LP distributions and for confidence and capital supply in the venture system. The 2018 Accel profile also acknowledges India's exit bottleneck; later sources forecast IPO counts. Invest in new ideas · Accel effect · 2021 crypto/IPO interview · 2025 IPO forecast
• [researcher inference · strong] A unicorn, announced IPO, offer price, or reported valuation is not realized fund return. The spoken corpus lacks vehicle ownership, dilution, primary/secondary splits, distributed proceeds, fees, carry, and fund cash flows. Internet-unicorn forecast · manufacturing-unicorn forecast

10. Venture methods applied to public and social systems—without merging the vehicles

• [investor-stated · moderate] ACT Grants shows a crisis-specific operating model: broad startup sourcing, specialist tracks, a collective IC, rapid grants to deployable teams, and reallocation as the pandemic changed. It is grantmaking and public-health capacity work, not Accel investing. ACT Grants interview
• [investor-stated · moderate] FCTC, MyGov, Sahyadri, and civic interviews extend the ecosystem thesis into policy, philanthropy, Bengaluru institutions, manufacturing, and public digital capacity. Those roles can create knowledge and networks, but public participation is not evidence of fund advantage or a conflict by itself. FCTC part 2 · MyGov · Sahyadri
• [researcher inference · strong] The missing control is a disclosed role map: which entity funded, recommended, selected, governed, or benefited; where portfolio interests overlapped; and when disclosure or recusal occurred. Master mentor · ACT Grants

Evolution across the record

Contradictions, counterevidence, and boundaries

1. Climate promise versus admitted underdelivery. The 2022 interview predicts India becoming a climate-tech epicenter. In 2025, Prakash directly says Indian climate tech had not lived up to potential; Mint frames sustainability, agritech, and EV as other underdelivering Accel theses. The later record is strategy-evolution evidence, not proof the new manufacturing thesis will work. Researcher inference · strong. Climate interview · Previous misses
2. No deep-tech ecosystem versus a manufacturing inflection. The 2023 critique identifies missing follow-on finance, university incentives, R&D, M&A, and LLM capability; 2025 says investor attitudes changed within six-to-nine months. These can be sequentially compatible, but the later claim needs measurable changes rather than event sentiment. Researcher inference · strong. Deep-tech gap · Manufacturing-unicorn forecast
3. Capital efficiency versus capital-compressed gestation. Early and post-boom sources favor doing more with less; manufacturing sources expect larger and faster rounds to compress development. The reconciliation is milestone-based financing, but Prakash does not publish a milestone/reserve/loss system. Researcher inference · strong. 2008 business plans · Manufacturing gestation
4. Founder choice versus investor power. The 2015 joint interview says founders cannot be forced; board, follow-on, preferred-security, runway, and distressed-company leverage can still create asymmetry. No adverse financing case tests the norm. Researcher inference · strong. Mobile-first interview
5. Intuition versus increased diligence. Guarded intuition and a prepared mind remain core, while the 2024 process takes nearly twice as long after governance failures. That is plausible learning, but elapsed time is not evidence of improved detection. Researcher inference · strong. Instinct · Investment pace and diligence
6. Broad sector mandates versus “not flavor of the season.” Interviews repeatedly list many fashionable sectors. The disciplined version is to treat lists as funnel areas and require a distinct company-level thesis; no public allocation record shows whether Accel did so. Researcher inference · strong. New ideas · Investment pace
7. Longevity democratization versus high-cost architecture. Biopeak's multi-omics, imaging, clinicians, and personalized roadmaps sit in tension with mass access. A later joint publisher summary attributed to Prakash and Nithin Kamath calls awareness, rather than affordability, the main adoption barrier, sharpening rather than resolving the missing price, reimbursement, and capacity evidence. The investing vehicle remains unclear. Researcher inference · strong. Biopeak investment · Medicine 4.0
8. Application-layer preference versus sovereign-model openness. June 2025 direct voice permits monetizable Indian sovereign models; a January 2026 organizer summary says foundation models are too capital intensive. Missing primary audio prevents a clean frontier-versus-targeted model distinction. Researcher inference · strong. IndiaAI allocation · SuperReturn summary
9. Public-market validation versus realized returns. Better margins, valuations, unicorns, IPO pipelines, and offer events recur as ecosystem validation. None reveals Accel ownership, dilution, proceeds, or distributed cash, much less Prakash-personal economics. Researcher inference · strong. Accel effect · IPO forecast
10. Mentor ideal versus downside record. The corpus is rich in favorable mentor language but thin on failed interventions, founder disputes, shutdown decisions, board resignations, or cases where advice harmed outcomes. The Kick-start profile and “previous misses” interview are valuable precisely because they break the success-only pattern. Researcher inference · strong. Master mentor · Kick-start fund · Previous misses
11. Civic access versus attribution/conflict controls. Policy and philanthropic networks can improve problem discovery and execution. They can also overlap with portfolio interests; no general disclosure, information-sharing, or recusal regime appears in the spoken record. This is an unanswered governance question, not an allegation of misconduct. Researcher inference · strong. FCTC part 2 · ACT Grants
12. Macro forecast versus investable edge. China+1, H-1B policy, urban clusters, government incentives, and capital geography recur. A directionally correct macro call can still fail to produce a priced, owned, realized investment. Researcher inference · strong. Global capital · H-1B · Manufacturing gestation

Biography, deals, vehicles, and attribution leads

Five sources to study first

The ranking is researcher inference · moderate and prioritizes decision utility rather than uniformly complete or direct evidence; completeness, directness, counterevidence, and source limitations remain source-specific.

1. Mint — Invest in new ideas, not flavour of the season — the strongest compact statement of prepared minds, decade-ahead themes, limited unilateral board control, 2021 excess, valuations, exits, and LP liquidity.
2. SeedToScale — Markets & Opportunities — the clearest reusable sourcing and picking sequence: predictable customer access and heavy lifting, unmet need, regional TAM, a tractable wedge, adjacent categories, and operating difficulty.
3. Mint — Master mentor — the best career and role narrative, including NetKraft failure/pivot, Erasmic formation, Accel combination, luck, mentoring, active-company load, and public/private boundaries.
4. FCTC part 2 — startups, politics and development — the broadest recent first-person integration of operator history, India venture evolution, manufacturing, AI, Flipkart recollection, and civic roles; caption limitations and attribution caveats remain explicit.
5. Mint — Accel rejigs India playbook after previous misses — the most important skeptical update: Prakash directly concedes climate-tech underdelivery, while Mint frames sustainability, agritech, and EV as other underdelivering Accel theses before presenting the new manufacturing thesis.

The 2008 business-plans interview is the best contemporaneous crisis/fund supplement; the Financial Express manufacturing Q&A the best current threshold and financing supplement; the Accel effect the best exit/governance counterweight.

Evidence still missing

• [researcher inference · strong] No complete personal deal ledger, pass denominator, IC memorandum set, partner vote/veto history, allocation schedule, ownership record, reserve history, fund cash flow, carry record, or audited person-level return attribution surfaced. Firm, team, vehicle, board, personal, and public-role claims therefore remain separated. Source map
• [researcher inference · strong] Flipkart, BlueStone, Infra.Market, MyGlamm, Captain Fresh, BookMyShow, QwikCilver, RentoMojo, and other named cases need contemporaneous memos, partner accounts, securities, vehicles, cap tables, board records, follow-ons, exits/write-offs, and cash flows before the talks become track-record evidence. Source map
• [researcher inference · strong] Diligence, mentoring, and board claims lack checklists, calendars, intervention logs, founder and co-investor testimony, failed interventions, conflict/recusal records, and matched outcomes. Longer diligence and repeated founder praise are not enough. Investment pace · Master mentor
• [researcher inference · strong] Manufacturing forecasts need cohort-level TRL at entry, certification/procurement milestones, capital raised, dilution, plant yield, revenue, gross margin, time to profitability, shutdowns, and realized exits. The three-to-five-year and unicorn windows remain open outcome checks. Manufacturing gestation · manufacturing-unicorn forecast
• [researcher inference · strong] AI claims need the 40-plus-company and 250-startup rosters, stage and vehicle data, compute/data rights, accuracy and safety, product deployment, monetization, ownership, survival, and returns. Service-workflow scale and voice access are hypotheses, not moats. IndiaAI allocation · AI voice-first
• [researcher inference · strong] ACT Grants and civic/policy roles need grant ledgers, selection minutes, implementation and impact evidence, related-party disclosures, information-sharing boundaries, and recusal practice. Nothing in the public record establishes a misconduct finding; the gap is governance visibility. ACT interview · FCTC part 2
• [researcher inference · strong] Biopeak needs a named investing vehicle, term sheet, ownership, governance, clinical endpoints, independent scientific review, pricing/access path, data rights, and IISc/portfolio conflict controls. The interview is interested advocacy, not medical validation. Biopeak investment
• [researcher inference · strong] The joint longevity appearances need speaker-separated transcripts plus clinical-validation evidence, model metrics, and vehicle attribution. Publisher summaries support explainability, traceability, human oversight, domain expertise, sound data, and the awareness-versus-affordability claim only jointly; they do not establish a Prakash-only doctrine. Medicine 4.0 · Data Science & AI
• [contemporaneous record · strong] Several media remain partial because YouTube blocked official captions, direct timed-text returned empty, publishers exposed only highlights, or no primary recording was released. Each note preserves the exact route and avoids invented quotes; obtaining creator captions or licensed media remains the cleanest closure path. COVID podcast · BlueStone deck · SuperReturn summary

Researcher inference · strong: the spoken corpus supports a coherent but bounded system—prepared judgment, unmet need and reachable customers, founder behavior, early-stage institutional capital, permissioned mentorship, capital discipline, IP-led manufacturing, application-layer AI, and ecosystem liquidity. It does not prove a complete personal process, person-level returns, causal board value, successful strategy changes, or adequate public/private conflict controls.

Prepared mind before non-consensus

• Rule — Build a dated view of a structural transition before it is fashionable, then treat intuition as a hypothesis to challenge with founder and market evidence rather than as permission to skip diligence. [researcher inference · strong] Talk synthesis, framework 1
• Evidence status — The prepared-mind and anti-fashion language is repeatedly investor-stated; application evidence is retrospective and selection-biased, so the reusable rule is [researcher inference · strong], not a verified source of excess return. Instinct and intuition note · Mobile-first Flipkart note
• Provenance and origin — The earliest surviving institutional form is Erasmic's 2008 multi-sector search rather than a personal slogan; Prakash later described conditioning and instinct in 2020 and explicitly urged investors in 2023 to pursue new ideas rather than consensus categories. [contemporaneous record · moderate] Erasmic strategy note · 2020 source · 2023 note Short origin quote: “At Accel, we have a set of prepared minds where we rejig and reassess spaces every four to five years.” Source note
• Mechanism and reasoning — A prior map reduces recognition time when technology, adoption, regulation, or talent changes; founder interaction then supplies disconfirming detail. The model works only if the view is timestamped, its causal mechanism is explicit, and subsequent evidence can reverse it. [researcher inference · strong] Writings chronology
• Historical and vehicle context — Early Indian venture markets had sparse adoption data, limited seed institutions, and small Erasmic cheques; later Accel vehicles had greater follow-on capacity. That evolution makes a prepared mind more scalable but also increases the cost of conviction errors and narrative escalation. [researcher inference · strong] Vehicle history
• When to use — Use it for emergent categories where historical datasets are thin but structural signals can be dated; do not use it to override unit economics, technical validation, conflicts, or a missing route to customers. [researcher inference · strong] Talks, contradiction boundary
• How to apply it — Write the transition, causal chain, leading indicators, disconfirmers, and next evidence date before meeting companies; ask what the founder knows that changes the map; kill or resize the thesis when the dated disconfirmers appear; record passes and conflicts so memory cannot retain only winners. [researcher inference · strong] Policybazaar miss
• Examples —
  • Retrospective application claim — QwikCilver: Balachandran and Prakash met the founders before stored value was widely understood; the later category account supports recognition and education work, but not the original memo or Prakash-only credit. Decision lab
  • Documented application — Captain Fresh: the team tracked Utham Gowda for a year and let his domain model refine its view of fragmented seafood supply. Accel essay
  • Researcher-applied illustration — Policybazaar/PB Fintech: Prakash says a Coverfox conflict blocked the first insurance wave and that Accel later obtained Acko exposure; garbled ASR, missing dates, and team attribution prevent a return counterfactual. Talk note, 19:35–20:38
• Evidence — The rule recurs across a 2020 joint interview, the 2015 Flipkart recollection, the 2023 new-ideas interview, Captain Fresh's jointly authored investment account, and the explicit insurance anti-portfolio. These are at least four claim placements across independent interview, talk, and firm-essay sources, but none publishes a complete accepted-and-rejected opportunity set. Talk synthesis · Captain Fresh note
• Limits, failure modes, and exception ledger — Intuition can become hindsight narration, sector fashion can masquerade as contrarianism, and a conflict can turn preparation into a documented pass. Flipkart's eventual scale was not anticipated, while climate optimism later met Prakash's underdelivery admission. [researcher inference · strong] Flipkart boundary · Climate miss
• Evolution and contradictions — The record moves from broad, instinct-heavy category discovery toward longer diligence and instrumented channel/manufacturing gates. The narrow reconciliation is intuition for where to look and evidence for whether to proceed; Prakash never states that full combined rule in one source. [researcher inference · moderate] Investment-pace note · Omnichannel framework
• Modern VC translation (as of 2026-08-20) — What transfers — timestamp the transition, causal chain, and falsifier before sourcing. What fails under changed 2026 conditions — the edge disappears when the theme is already consensus or the dated cost, policy, or adoption premise reverses. Observable applicability signals — versioned assumptions, disconfirming data, and a founder-caused update. Most likely misuse — relabeling a fashionable AI, climate, or manufacturing theme as contrarian after a winner appears. Analyst-created hypothetical — write an industrial-automation memo before meetings and resize the theme when its adoption gate misses. [researcher inference · strong] Prepared-minds source · Climate counterevidence

Reachable customer and bounded bottleneck

• Rule — Start with a specific unmet need and a customer the company can reach predictably; bound the initial market, then require evidence that the wedge can expand into adjacent demand. [investor-stated · moderate] Markets and opportunities note
• Evidence status — The sequence is directly stated in the spoken corpus and independently echoed by Erasmic's customer-validation screen and investment essays; the claim that it predicts returns is [researcher inference · provisional] because the rejected-market denominator is absent. Funding philosophy note
• Provenance and origin — Erasmic's captured 2008 funding philosophy preferred operating companies with customer validation and a credible India revenue path; the later markets talk makes customer access, unmet need, regional TAM, and adjacency explicit. [contemporaneous record · strong] Common Crawl record · Markets note Short origin quote: “India is not homogenous or singular when compared to other countries.” Source note
• Mechanism and reasoning — Reachable customers shorten learning cycles and expose willingness to pay; a bounded wedge makes adoption, contribution, and retention falsifiable; adjacency supplies venture scale only after the wedge works. [researcher inference · strong] Talk synthesis, framework 2
• Historical and vehicle context — India's early adoption, payments, logistics, and disposable-income constraints made broad top-down TAM especially unreliable. Small early cheques could finance customer proof, whereas later fund scale required credible expansion beyond a niche. [researcher inference · strong] Erasmic Indian landscape · Accel Effect
• When to use — Use for seed and Series A markets with uncertain adoption or geography-specific behavior; do not use a tractable wedge when it cannot support ownership economics, when expansion has a different buyer, or when customer access depends on unsustainable discounts. [researcher inference · strong] Customer-engagement note
• How to apply it — Name the buyer, unmet job, current workaround, acquisition route, purchase trigger, regional denominator, wedge economics, and adjacency assumptions; require paid or behaviorally costly proof; stop when access, repeat, margin, or adjacency conversion misses a pre-registered threshold. [researcher inference · strong] Decision playbook
• Examples —
  • Documented application — Captain Fresh: exact buyer specifications and variable seafood supply define a coordination bottleneck; the authors still disclose no margin or working-capital proof. Investment essay
  • Documented application — Teabox: direct access to global tea buyers bypassed slow intermediaries, but logistics and technology were visible constraints; FY24 reporting suggests a small profitable company rather than a realized venture win. Seed announcement · Outcome control
  • Researcher-applied illustration — BabyOye: specialist selection and online convenience were plausible, but inventory, fulfillment, repeat, and CAC were not publicly documented decision gates; the later acquisition was reported as a likely major VC write-off. Decision lab
• Evidence — Independent support comes from the 2008 Erasmic page, the markets talk, a Captain Fresh investment essay, a Teabox period announcement, and BabyOye's failure record. Four-plus placements support the screen, while vehicle ownership and comparative success rates remain missing. Writings synthesis · Talk synthesis
• Limits, failure modes, and exception ledger — A vivid problem can hide a small market, low willingness to pay, poor margins, regulation, or a different adjacency buyer. Teabox is the ordinary-outcome control; BabyOye shows that customer convenience does not solve physical economics. [researcher inference · strong] Teabox and BabyOye controls
• Evolution and contradictions — The model evolves from customer validation and India revenue in 2008 to direct engagement in 2016 and measured online-to-offline adjacency in 2024. Broad forecasts about the next billion users remain less instrumented than the operational rule. [researcher inference · strong] Internet forecast · Omnichannel report
• Modern VC translation (as of 2026-08-20) — What transfers — name the buyer, unmet job, reachable channel, bounded wedge, and adjacency. What fails under changed 2026 conditions — AI demo usage or a large language population does not transfer when paid repeat, distribution ownership, or inference-adjusted contribution is absent. Observable applicability signals — paid conversion, repeat workflow completion, cohort margin, and adjacency conversion. Most likely misuse — treating India’s population or downloads as TAM. Analyst-created hypothetical — fund a Hindi voice field-workflow only after one paid role repeats, then gate the next language on cohort evidence. [researcher inference · strong] Market framing · Voice-first boundary

Founder learning under ambiguity

• Rule — Prefer founders who have earned specific insight by operating inside a hard system, can teach the investor something falsifiable, make costly customer-centered choices, and update without abandoning responsibility. [researcher inference · strong] Writings decision playbook
• Evidence status — Domain obsession, customer work, curiosity, decisiveness, compatibility, and resilience are investor-stated across independent sources; their predictive value is [researcher inference · provisional] because no matched founder-rejection or team-failure dataset is public. Talk synthesis, framework 3
• Provenance and origin — Erasmic's 2008 screen required working compatibility and a domain where managers could add value; later company essays make founder-specific learning more explicit. [contemporaneous record · strong] Funding philosophy · Captain Fresh essay Short origin quote: “We have always looked for founders who are passionate about solving hard problems in seemingly under-the-radar sectors.” Source note
• Mechanism and reasoning — Lived exposure generates non-obvious process, customer, and regulation knowledge; falsifiable teaching reveals whether insight is causal rather than charismatic; costly decisions show priorities under constraint; updating reduces attachment to the first model. [researcher inference · strong] Accel profile
• Historical and vehicle context — Operator-investors in Erasmic could recognize execution detail and offer labor, but that also invited affinity bias. Accel's wider platform increased specialist access while making individual sourcing and intervention attribution harder. [researcher inference · strong] Erasmic team note · People map
• When to use — Use at seed and in opaque physical or regulated systems where learning speed matters; do not substitute personality adjectives for references, technical proof, governance, or economics, and do not infer a founder trait from a later successful outcome. [researcher inference · strong] Talks limits
• How to apply it — Ask for one operating insight competitors miss, three decisions changed by evidence, a painful customer problem personally observed, a reference who saw the change, and the unresolved expert gap; kill the deal when the story cannot be reproduced or contrary evidence is rationalized away. [researcher inference · strong] Modern founder-learning translation
• Examples —
  • Documented application — Captain Fresh: the authors say Gowda's domain obsession and ability to explain the system built conviction during a year of tracking. Accel essay
  • Retrospective application claim — BlueStone: Prakash emphasizes first-principles curiosity, local jewelry knowledge, endurance, and willingness to change the model; the account was written at IPO and needs hindsight discounting. IPO reflection
  • Documented application — MyGlamm/Good Glamm: the 2021 essay treated a near-failure pivot and decisive channel choices as resilience; later lender-driven breakup shows resilience did not assure integration, solvency, or governance. Original thesis · Failure lab
• Evidence — Four-plus placements come from an archived 2008 team screen, the 2015 Flipkart recollection, Captain Fresh, BlueStone, MyGlamm, and a later failure record. Joint authorship and retrospective framing are retained at each case. Flipkart note · Writings synthesis
• Limits, failure modes, and exception ledger — Passion, humility, obsession, and resilience are narratively elastic; compatible founders can still make poor financing or governance choices; investor operating similarity may create affinity bias. Good Glamm is the strongest negative control, and no public rejected-founder or post-investment team-change denominator was located. [researcher inference · strong] Good Glamm postmortem
• Evolution and contradictions — Early compatibility and customer validation broaden into learning, curiosity, resilience, and mission; after governance failures Accel reportedly lengthened diligence. Longer diligence is a process change, not evidence the founder screen became more accurate. [researcher inference · strong] Investment-pace interview
• Modern VC translation (as of 2026-08-20) — What transfers — test earned domain insight, falsifiable teaching, and evidence-driven updating. What fails under changed 2026 conditions — founder polish, generated prototypes, intensity, or mission language cannot substitute for references and reproduced workflow knowledge. Observable applicability signals — specific changed decisions, failed experiments, customer references, and an admitted expert gap. Most likely misuse — converting charisma or resilience into a hindsight trait. Analyst-created hypothetical — ask an industrial founder to reproduce one yield lesson and show the decision it reversed before scoring learning. [researcher inference · moderate] Captain Fresh founder evidence · Good Glamm boundary

Evidence and capital efficiency before scale

• Rule — Finance growth only after customer evidence and a credible first-rupee profit mechanism show how capital converts into durable operating progress; burn is acceptable only against pre-specified evidence and financing fit. [researcher inference · strong] Profit-mechanism note
• Evidence status — Customer validation, capital efficiency, sustaining revenue, profitability, and financing fit recur from 2008 to 2026; company-level enforcement and superior outcomes remain [researcher inference · provisional]. Talk synthesis, framework 6
• Provenance and origin — Erasmic's 2008 doctrine preferred differentiated, comparatively low-capital companies with a credible revenue path; the 2008 downturn interview sharpened cash conservation, and later talks require a visible route to profit. [contemporaneous record · strong] Funding philosophy · Business-plans note Short origin quote: “Staged infusion of capital is the most efficient and beneficial to all parties in the long run”. Source note
• Mechanism and reasoning — Early evidence reduces financing dependence, while a capital-to-milestone map exposes whether growth creates retention, margin, or defensibility. Matching equity, debt, working capital, and asset finance to cash-flow risk prevents a liquidity instrument from hiding a business-model problem. [researcher inference · strong] Venture-debt note
• Historical and vehicle context — Sparse 2008 seed capital made efficiency a survival constraint; later larger funds and abundant rounds could finance longer gestation but also escalation. Public records disclose neither reserve policy nor loss budget, so vehicle incentives cannot be assumed benign. [researcher inference · strong] Vehicle comparison
• When to use — Use before every initial and follow-on decision, especially commerce, logistics, hardware, and roll-ups; do not require premature accounting profit when a measured investment creates durable density, quality, certification, or network value. [researcher inference · strong] Accel Effect note
• How to apply it — Define contribution after the real bottleneck, repeat/retention, working-capital cycle, cash runway, milestone cost, downside financing, and a no-next-round plan; match the security to the risk; stop when capital no longer narrows uncertainty or unit economics fail the dated threshold. [researcher inference · strong] Failure labs
• Examples —
  • Retrospective application claim — BlueStone: frugality, unit economics, catalog expansion, and capability building are presented as surviving a capital-scarce middle; reserve decisions and causality are missing. IPO reflection
  • Researcher-applied illustration — UrbanTouch: a reproduced tribunal record gives ₹58.75 million invested and ₹35.02 million sale consideration, a bounded roughly 0.596x observed ratio before unobserved cash flows. Decision lab
  • Researcher-applied illustration — Good Glamm: a reported flat bridge, debt/default, delayed obligations, and CIRP show that growth capital and acquisitions can postpone rather than resolve cash-generation risk. NCLT-backed lab
• Evidence — Independent support comes from the 2008 Erasmic screen, the downturn interview, profitability talks in 2022 and 2026, BlueStone's retrospective, and two legal/failure controls. Profitability note · 2026 note · Failure postmortems
• Limits, failure modes, and exception ledger — Short-term profitability can reflect underinvestment, layoffs, or asset sales; efficiency can underfund a genuine network or certification curve; venture debt can move risk to covenants and refinancing. The public record lacks cohort-level burn, reserves, and write-offs. [researcher inference · strong] Venture-debt boundary
• Evolution and contradictions — The rule persists, but larger vehicles and manufacturing optimism make it more conditional: Prakash argues capital can compress gestation while later demanding profitability and public-market discipline. Money cannot compress regulatory validation or physics. [researcher inference · strong] Manufacturing-gestation note
• Modern VC translation (as of 2026-08-20) — What transfers — measure capital per uncertainty retired and match the financing instrument to the cash-flow risk. What fails under changed 2026 conditions — abundant equity or venture debt does not repair weak repeat, contribution, covenants, or a missing no-next-round plan. Observable applicability signals — retention, contribution after the real bottleneck, runway, cash conversion, and milestone cost. Most likely misuse — calling underinvestment or a small non-venture outcome efficient. Analyst-created hypothetical — release a commerce follow-on only after a cohort clears repeat and post-returns contribution gates. [researcher inference · strong] Venture-debt limits · 2026 profitability screen

Fragmentation to standardized coordination

• Rule — Invest where technology plus operations can convert fragmented, variable supply and demand into a repeatable service with measurable quality, availability, and working-capital economics. [researcher inference · strong] Writings synthesis
• Evidence status — The mechanism is jointly investor-stated across the Golden Age and Captain Fresh accounts and appears in several company cases; its outcome efficacy is [researcher inference · provisional] because failures and ordinary outcomes are incompletely observed. Captain Fresh note
• Provenance and origin — Erasmic described cross-border operating gaps in 2008; the 2020 Golden Age essay explicitly elevated B2B aggregation and supply-chain digitization, and the 2021 Captain Fresh essay supplied the most detailed operating example. [contemporaneous record · strong] Golden Age · Captain Fresh Short origin quote: “Before we met Utham, we only had a vague idea of how fragmented the seafood supply chain is” Source note
• Mechanism and reasoning — A coordinator can pool supply, translate buyer specifications, forecast demand, reduce search and failure costs, and create data from repeated workflows. The moat exists only if service reliability and contribution improve faster than inventory, credit, spoilage, and operational complexity. [researcher inference · strong] Infra.Market note
• Historical and vehicle context — Indian physical markets combined fragmented suppliers, uneven logistics, and fast digitization; larger Accel vehicles could fund working capital and physical capability that a small seed fund could not. This creates both opportunity and capital-intensity risk. [researcher inference · strong] Golden Age note · Vehicle boundary
• When to use — Use where fragmentation causes observable mismatch, delay, waste, quality variance, or financing friction; do not use it when incumbents coordinate cheaply, take rates cannot cover operations, or the platform bears risks customers will not pay to remove. [researcher inference · strong] Captain Fresh risk boundary
• How to apply it — Map every supplier, buyer, handoff, quality failure, inventory day, credit exposure, and data loop; identify the minimum owned operation; require service-level improvement, repeat demand, positive post-loss contribution, and declining working-capital intensity before geographic expansion. [researcher inference · strong] Deep Rooted lab
• Examples —
  • Documented application — Captain Fresh: heterogeneous seafood supply, fragile logistics, and buyer-specific requirements motivate a software-and-operations coordinator; no public economics validate the thesis. Accel essay
  • Researcher-applied illustration — Infra.Market: Prakash describes technology, aggregation, and later owned plants; the evolution shows that coordination can become asset ownership, while private economics remain unavailable. Talk note
  • Researcher-applied illustration — Deep Rooted: greenhouse aggregation and demand-backed perishables evolved into a capital-intensive channel pivot; FY24 losses preceded a reported operating shutdown. Decision lab
• Evidence — Four-plus placements span the 2020 ecosystem essay, a 2021 investment essay, an Infra.Market interview, Deep Rooted's period financing and warning sources, and its later failure report. Global AgInvesting · Deep Rooted warnings
• Limits, failure modes, and exception ledger — Coordination can hide structurally low margins, working-capital drag, spoilage, credit losses, and local density requirements. Deep Rooted is the strongest negative control; Teabox is an ordinary small outcome; Captain Fresh and Infra.Market lack public return evidence. [researcher inference · strong] Ordinary and failure controls
• Evolution and contradictions — The corpus begins with asset-light preferences but later celebrates full-stack operations and owned plants. The coherent rule is not “asset light”; own only what measurably improves coordination and economics. [researcher inference · strong] Accel Effect · Writings tension 3
• Modern VC translation (as of 2026-08-20) — What transfers — map every fragmented handoff and test whether coordination improves a paid service level. What fails under changed 2026 conditions — software does not create a moat when inventory, credit, spoilage, or incumbent coordination absorbs the gain. Observable applicability signals — fill rate, rejection, repeat, loss-adjusted contribution, inventory days, and cash conversion. Most likely misuse — applying SaaS valuation to a financed operating balance sheet. Analyst-created hypothetical — pilot one industrial-input corridor and expand only after service improves while working-capital intensity falls. [researcher inference · strong] Captain Fresh mechanism · Deep Rooted adverse case

Minimum necessary full stack

• Rule — Own the minimum set of software, supply, logistics, service, or physical layers required to make trust, quality, and availability reproducible; outsource layers that do not create measurable control or learning. [researcher inference · strong] Writings operating rules
• Evidence status — Prakash and co-authors repeatedly describe full-stack responses to Indian operating gaps, but “minimum necessary” is the researcher’s reconciliation of early asset-light preferences with later owned capability. Outcome causality is [researcher inference · provisional]. Golden Age · Talk synthesis
• Provenance and origin — Erasmic’s 2008 screen favored comparatively low-capital businesses; by 2018–2021, the Accel corpus explicitly argued that Indian startups sometimes had to build missing ecosystem layers. [contemporaneous record · strong] Erasmic philosophy note · Accel Effect Short origin quote: “the business has to be tech-enabled in some way”. Source note
• Mechanism and reasoning — Owning a bottleneck can reduce variance, shorten feedback, guarantee service, and create proprietary data; every added layer also increases fixed cost, working capital, execution surfaces, and governance burden. The economic test is incremental control per unit of capital. [researcher inference · strong] Captain Fresh note
• Historical and vehicle context — Missing logistics, payments, quality assurance, and trusted retail made the Indian “stack” wider than a pure software analogue. Larger Accel funds could finance that width, but public records do not disclose reserves or layer-level returns. [researcher inference · strong] Vehicle boundary
• When to use — Use when a named external failure prevents customer adoption or quality; do not integrate merely to increase narrative scope, reported gross merchandise value, or valuation, and do not own a layer whose failure risk exceeds the control it provides. [researcher inference · strong] Capital-efficiency framework
• How to apply it — Map the customer promise and each failure point; price build, partner, and contract alternatives; pilot one owned layer; require a pre-set improvement in fill rate, defects, repeat, margin, or cash conversion; reverse or stop if the layer misses. [researcher inference · strong] Historical operating rule
• Examples —
  • Retrospective application claim — BlueStone: an online jeweler added physical stores and manufacturing/fulfillment capabilities to solve trust and experience; the IPO reflection does not isolate each layer’s return. IPO reflection
  • Documented application — Captain Fresh: software, quality control, logistics, and trade execution answer variable seafood supply; private margins and capital intensity remain undisclosed. Accel essay
  • Documented application — MyGlamm/Good Glamm: content, commerce, brand acquisition, and offline distribution broadened the stack, but later debt distress and CIRP show scope did not guarantee solvency or integration. Original thesis · Failure lab
• Evidence — Independent placements include the archived Erasmic screen, the Accel Effect interview, Golden Age essay, BlueStone and Captain Fresh firm essays, and Good Glamm’s adverse record. They support the tension and cases, not a published layer-level underwriting formula. Writings synthesis · Failure postmortem
• Limits, failure modes, and exception ledger — Vertical integration can disguise weak product-market fit, cross-subsidize loss-making layers, and create inventory, credit, compliance, or founder-bandwidth risk. Good Glamm is the clearest negative control; BlueStone’s survivor account is hindsight-prone. [researcher inference · strong]
• Evolution and contradictions — The record moves from low-capital preference to acceptance of full-stack operating companies. The bounded synthesis is to integrate only the causal bottleneck, instrument it, and remain willing to remove it; Prakash does not state that complete rule verbatim. [researcher inference · strong] Writings contradictions
• Modern VC translation (as of 2026-08-20) — What transfers — own only the layer that causally improves trust, quality, availability, or learning. What fails under changed 2026 conditions — integration loses its rationale when a current provider, API, contract, or standard meets the service threshold more cheaply. Observable applicability signals — layer-level defects, fill rate, repeat, contribution, cash need, and a credible outsource comparison. Most likely misuse — calling scope or operational complexity a moat. Analyst-created hypothetical — an AI clinic owns quality review but partners for logistics unless the partner misses a pre-registered service gate. [researcher inference · strong] Full-stack history · Good Glamm scope boundary

Density-gated omnichannel expansion

• Rule — Let online demand reveal local density, pilot a small physical format in the strongest catchment, measure incremental economics and cannibalization, and scale stores only when repeatable gates clear. [researcher inference · strong] Omnichannel report
• Evidence status — The Accel–Fireside framework directly states the sequence and metrics, while company examples are retrospective and jointly attributed. No Prakash-specific store approval ledger or matched failed-pilot sample is public. [contemporaneous record · strong] Toolkit note
• Provenance and origin — Earlier consumer essays emphasized mobile-first reach and direct engagement; the 2024 report formalized online brand-building, order-density mapping, pilot stores, catchments, and expansion metrics. [contemporaneous record · strong] Mobile-first note · Report note Short origin quote: “Early signs of a consumer franchise being built online paves the path for planning offline expansion.” Source note
• Mechanism and reasoning — Digital demand lowers discovery cost and maps geographic intent; a store can add trust, trial, convenience, and service. Catchment-level incrementality distinguishes new demand from customers merely shifting channels. [researcher inference · strong] Omnichannel talk note
• Historical and vehicle context — India’s fragmented retail, rising internet access, and uneven city density made digital-first entry attractive, while later consumer funds could finance stores. Physical rollout changes capital, lease, inventory, and governance exposure. [researcher inference · strong] Golden Age note
• When to use — Use for categories where touch, fit, trust, service, or immediacy constrains online conversion; do not use stores to conceal weak online retention, acquire vanity footprint, or copy a winner across low-density catchments. [researcher inference · strong] Report
• How to apply it — Geocode orders and searches; select the strongest catchment; pre-register store capex, payback, incremental revenue, repeat, gross margin, inventory turns, and cannibalization; pilot one format; compare with control catchments; stop or redesign before rollout if gates fail. [researcher inference · strong] Toolkit note
• Examples —
  • Retrospective application claim — BlueStone: the company moved from online jewelry to stores and later IPO; the source describes endurance and trust but publishes no controlled catchment analysis. Accel IPO reflection
  • Documented application — Lenskart: the joint report uses digitally informed physical expansion as a category exemplar; it is not a disclosed Prakash investment memo. Omnichannel report
  • Researcher-applied illustration — Deep Rooted: a demand-backed perishables model pivoted toward a consumer channel, yet losses and a reported operational shutdown show that channel expansion cannot repair density, margin, or working-capital failure by itself. Decision lab
• Evidence — Four-plus placements come from the full report, two contemporaneous report/talk notes, BlueStone’s firm reflection, and Deep Rooted’s financing and shutdown evidence. The evidence supports a measurement framework, not causal proof of store-led returns. Report note · Failure postmortem
• Limits, failure modes, and exception ledger — Geocoded orders can reflect marketing allocation rather than latent demand; attribution windows, privacy limits, leases, inventory transfers, and channel cannibalization can distort payback. Deep Rooted is a channel-pivot negative control, not a store-format test. [researcher inference · strong]
• Evolution and contradictions — Digital-first conviction evolved into a measured online-plus-offline system. The later framework corrects any simplistic “internet replaces stores” reading, but does not establish that every digital brand should become omnichannel. [researcher inference · strong] Writings evolution
• Modern VC translation (as of 2026-08-20) — What transfers — use online order density to select a physical pilot and measure catchment-level incrementality. What fails under changed 2026 conditions — aggregate growth is uninformative when attribution, leases, inventory transfers, privacy limits, or cannibalization prevent a clean local comparison. Observable applicability signals — control catchments, incremental repeat, gross margin, inventory turns, capex payback, and cannibalization. Most likely misuse — rolling out stores from brand heat or blended revenue. Analyst-created hypothetical — open one diagnostic center in the densest catchment and stop if holdout-adjusted payback misses. [researcher inference · strong] 2024 framework · Channel-pivot boundary

Patient capital with milestone gates

• Rule — Extend time and follow-on capital only while a company is retiring named technical, customer, regulatory, or channel uncertainties; pair patience with dated milestones, reserve limits, and an explicit stop or restructure rule. [researcher inference · strong] Capital-gestation note
• Evidence status — Prakash explicitly argues for patient capital and greater pools for long-gestation sectors, while company histories show endurance. The milestone-gate and reserve-limit synthesis is [researcher inference · provisional] because no public Accel reserve policy or Prakash IC record was located. Talk synthesis
• Provenance and origin — Erasmic proposed small initial checks plus hands-on help; Accel’s later vehicles and Prakash’s manufacturing comments add follow-on capacity and long-duration ambition. [contemporaneous record · moderate] Erasmic overview · Manufacturing note Short origin quote: “We typically engage with companies that have at least a TRL-6 (technology readiness level) product.” Source note
• Mechanism and reasoning — Some uncertainty resolves only through iteration, certification, density, or behavior change; sufficient runway preserves option value. Milestones prevent elapsed time and sunk capital from being mistaken for progress, and stop rules protect the portfolio from narrative escalation. [researcher inference · strong] BlueStone reflection
• Historical and vehicle context — Small Erasmic vehicles constrained initial exposure; later Accel India funds could support multiple rounds and sectors with longer qualification. Larger reserves also create incentive to defend ownership or continue weak bets, but public cash-flow data are absent. [researcher inference · strong] Vehicle comparison
• When to use — Use when the uncertainty has a measurable resolution path and the resulting market could fit the vehicle; do not confuse regulatory waiting, repeated pivots, or financing need with learning, and do not promise that capital can compress physics. [researcher inference · strong]
• How to apply it — Define the next falsifiable proof, time and cash required, reserve cap, external dependency, downside financing, governance owner, and stop/restructure choice before each tranche; compare progress with the no-follow-on counterfactual at every gate. [researcher inference · strong] Decision labs
• Examples —
  • Retrospective application claim — BlueStone: Accel describes a long hold through a capital-scarce middle and eventual IPO; no tranche-level milestones or counterfactual are published. IPO reflection
  • Researcher-applied illustration — Deep Rooted: multiple rounds followed greenhouse and channel pivots before losses and reported shutdown; public sources do not reveal who approved follow-ons or which gates failed. Decision lab
  • Researcher-applied illustration — UrbanTouch: the reproduced tribunal record implies roughly 0.596x on disclosed investment versus sale consideration before unknown cash flows; it demonstrates a stop, not whether the timing was optimal. Decision lab
• Evidence — Independent support comes from Erasmic’s archived check-size model, fund-history interviews, the BlueStone reflection, a manufacturing interview, and the Deep Rooted and UrbanTouch adverse records. These do not disclose capital calls, reserves, internal gates, or Prakash’s vote. Fund interview · Adverse postmortems
• Limits, failure modes, and exception ledger — Patient capital can subsidize weak economics, postpone governance action, and crowd out new options; premature stops can destroy learning or future value. Attempted routes included Erasmic/Accel archived pages, fund interviews, Form D/CIMA and Rockefeller 990 references, public-pension searches, and company/legal records; no LPAs, IC minutes, reserve schedules, capital calls, distributions, or complete follow-on decisions were found. Below target for documented application.
• Evolution and contradictions — The philosophy expands from low-capital seeds to institutionally supported, longer-gestation companies, while later profitability language adds a corrective. Prakash’s claim that more capital can compress gestation is bounded by his own climate underdelivery admission and by non-financeable scientific or regulatory time. [researcher inference · strong] Climate miss
• Modern VC translation (as of 2026-08-20) — What transfers — extend time only while named technical, customer, regulatory, or channel uncertainty is retiring. What fails under changed 2026 conditions — larger rounds cannot compress physics, certification, procurement, or repeated pivots that create no new evidence. Observable applicability signals — dated proof point, cash to proof, reserve cap, external dependency, and stop or restructure decision. Most likely misuse — describing an uncapped rescue as patience. Analyst-created hypothetical — tranche a robotics round at qualification and repeat-yield gates rather than elapsed quarters. [researcher inference · strong] Capital-compression claim and limits · Deep Rooted outcome

Permissioned mentoring with capacity limits

• Rule — Offer operating help only with founder consent, a named bottleneck, a capable owner, and an explicit handoff; limit active commitments so mentoring does not become shadow management or superficial access. [researcher inference · strong] Mentor talk note
• Evidence status — Permission, willingness, compatibility, operating contribution, and capacity recur in Prakash’s mentoring talk and archived Erasmic pages. Overall efficacy is [researcher inference · provisional]. Claimed interventions are documented, but incremental company outcomes and Prakash-only attribution are not. Talk synthesis
• Provenance and origin — Erasmic marketed management participation and practical assistance; the later mentoring talk makes recipient willingness and finite mentor capacity explicit, while Prakash’s board self-critique supplies a governance boundary. [contemporaneous record · strong] Value-add archive · Governance note Short origin quote: “In my mind, it is being a guide and mentor” Source note
• Mechanism and reasoning — Consent improves information flow and preserves founder agency; a named problem and owner make help actionable; a handoff prevents dependency; a capacity limit protects attention. Board authority changes the relationship and requires formal accountability rather than mentoring language. [researcher inference · strong]
• Historical and vehicle context — A small operator-led Erasmic portfolio could plausibly provide intensive help; Accel’s larger platform offered deeper recruiting and network capacity but diluted personal attribution and increased portfolio-load risk. [researcher inference · strong] Erasmic value-add note · People map
• When to use — Use when a founder requests help on a solvable hiring, customer, operating, or governance bottleneck; decline when incentives conflict, expertise is absent, the board must exercise formal oversight, or attention cannot be sustained. [researcher inference · strong]
• How to apply it — Obtain permission; write the problem, owner, decision right, deliverable, time box, and exit; disclose conflicts; measure whether the bottleneck improved; hand back responsibility; reduce commitments when the capacity ledger exceeds a pre-set limit. [researcher inference · strong] Erasmic positioning note
• Examples —
  • Researcher-applied illustration — Erasmic portfolio: archived pages describe help with recruiting, customers, technology, and management; these are contemporaneous marketing claims, not independently measured effects. Value-add archive
  • Retrospective application claim — BabyOye: a founder account credits Prakash with operating help, yet later reports imply a major write-off; useful help does not establish investment success or causal rescue. Decision lab
  • Researcher-applied illustration — BlueStone: long board association and an eventual IPO are documented, but no source identifies which decisions Prakash personally changed. IPO reflection
• Evidence — Four-plus placements span three archived Erasmic pages, the full mentoring talk, the board-independence talk, BabyOye’s founder account, and BlueStone’s firm essay. They support a process ethic, not a causal return claim. Mentor note · Funding philosophy
• Limits, failure modes, and exception ledger — Permission can be performative when the investor controls capital or a board seat; advice can blur accountability, create key-person dependency, or exceed sector expertise. BabyOye is the negative outcome control, and Prakash’s own capacity estimate of roughly 15–18 active relationships is an asserted ceiling rather than audited workload. [researcher inference · strong] Attempted routes: the complete Mint mentoring profile, archived Erasmic overview/value-add/positioning/philosophy pages, the BabyOye founder account, BlueStone and Accel essays, the live Accel profile, talk synthesis, and investment/legal records. Results: general claimed help and a 15–18 capacity statement, but no consent log, intervention register, hours, named owner/handoff, matched outcome set, or board minutes; demonstrated mentoring efficacy is below target. Mentor note · Erasmic value-add note · Investment retrieval ledger
• Evolution and contradictions — Hands-on operator help becomes a more explicit permission-and-capacity model, while governance commentary distinguishes mentorship from board duty. The residual contradiction is structural power: consent between investor and founder is never fully symmetrical. [researcher inference · strong] Governance note
• Modern VC translation (as of 2026-08-20) — What transfers — require consent, a named bottleneck, an owner, a time box, and a handoff. What fails under changed 2026 conditions — larger portfolios and formal board duties make informal help unsafe when power, conflicts, hours, or accountability are unrecorded. Observable applicability signals — founder request, scoped deliverable, logged time, outcome, conflict check, and handoff date. Most likely misuse — counting introductions or investor intervention as causal value-add. Analyst-created hypothetical — time-box a recruiting search and return the decision to the founder after a measured slate is delivered. [researcher inference · strong] Mentoring account · Governance boundary

Fund-scale portfolio and follow-on fit

• Rule — Match opportunity size, entry check, target ownership, follow-on need, partner time, and plausible exit value to the vehicle before investing; a larger fund increases support capacity but also changes incentives and the minimum outcome that matters. [researcher inference · strong] Vehicle analysis
• Evidence status — Check-size evolution, early-stage positioning, and greater follow-on capacity are documented at firm level. Ownership targets, reserve ratios, concentration, partner attribution, and realized vehicle economics are not public, so tested portfolio construction is [researcher inference · insufficient public record] and below target. Talk synthesis
• Provenance and origin — Erasmic’s archived overview proposed four to six initial investments of roughly $100,000–$500,000; later Accel India funds expanded institutional and follow-on capacity. [contemporaneous record · moderate] Overview archive · 2012 fund interview Short origin quote: “picking a set of 4-6 companies” Source note
• Mechanism and reasoning — Fund size determines how much ownership and return each company must supply; follow-on reserves preserve winners and finance proof, while partner time constrains operating help. Misfit encourages overcapitalization, premature market expansion, or attention to companies too small to affect the fund. [researcher inference · strong]
• Historical and vehicle context — Prakash’s record spans Erasmic, Accel India vehicles, and separate family/philanthropic structures. Firm investments cannot be attributed automatically to him, and family or grant capital cannot be blended into venture returns. [researcher inference · strong] Attribution rules
• When to use — Use before initial checks, follow-ons, opportunity-fund transfers, and any cross-vehicle decision; reject or route elsewhere when the company’s capital path, exit range, duration, governance load, or mandate does not fit. [researcher inference · strong]
• How to apply it — Model entry and diluted ownership, total reserve requirement, partner hours, loss ratio, exit distribution, and contribution to fund return; define cross-fund allocation and conflict rules; compare with the next-best use of reserves at every follow-on. [researcher inference · strong]
• Examples —
  • Researcher-applied illustration — Erasmic: the archived page proposed small checks into a handful of early companies paired with operating help; it does not prove actual portfolio construction or realized returns. Overview note
  • Researcher-applied illustration — BlueStone: sustained association through IPO fits follow-on capacity, but cash invested, proceeds, ownership dilution, and fund allocation are not fully public. Investment economics
  • Researcher-applied illustration — Deep Rooted: repeated financing and an operating shutdown raise reserve-allocation questions; sources do not identify Prakash’s personal approvals, vehicle exposure, or stop rule. Failure lab
• Evidence — Four-plus placements include the archived Erasmic overview and philosophy, 2012 and 2024 fund interviews, BlueStone’s IPO record, and Deep Rooted’s financing/outcome record. These establish vehicle evolution and examples, not a validated allocation model. Investment-pace note · Vehicle comparison
• Limits, failure modes, and exception ledger — Fund math can rationalize overownership, starve experimental checks, or hide partner-level conflicts. Attempted routes included archived Erasmic/Accel pages, Form D and CIMA references, Rockefeller 990 materials, public-pension minutes, fund interviews, company filings, and legal records; no LPAs, IC/allocation minutes, reserve schedules, capital calls, distributions, company-by-company fund mapping, or complete performance denominator were found.
• Evolution and contradictions — The platform grew from concentrated operator-led seed investing into a larger early-stage institution. Greater capacity supports patience and specialization but weakens simple claims of personal causality and raises the hurdle for small outcomes such as Teabox. [researcher inference · strong] Ordinary-outcome control
• Modern VC translation (as of 2026-08-20) — What transfers — match entry, diluted ownership, reserves, partner time, duration, and plausible exit contribution to the vehicle. What fails under changed 2026 conditions — a larger platform does not solve a deal whose total capital path, small outcome, cross-fund allocation, or governance load cannot matter economically. Observable applicability signals — modeled ownership, reserve ratio, loss budget, partner-capacity ledger, and allocation rule. Most likely misuse — treating assets under management or financing rounds as performance. Analyst-created hypothetical — route a capital-light niche company to a smaller vehicle when even a strong exit cannot move the flagship fund. [researcher inference · strong] Fund-size constraints · Erasmic design

IP and qualification led manufacturing

• Rule — Back manufacturing only when a team owns a defensible product, process, or equipment advantage and can traverse prototype, qualification, repeat production, yield, and global customer relevance. [researcher inference · strong] Manufacturing-strategy note
• Evidence status — Prakash directly states the IP-and-engineering screen and the climate-tech underdelivery; the product/process/equipment taxonomy is firm-level, while qualification gates are [researcher inference · provisional]. No complete manufacturing deal or outcome cohort is public. Mint interview
• Provenance and origin — After broad sustainability optimism and a 2023 admission that India lacked a real deep-tech ecosystem, the 2025 playbook narrowed toward proprietary engineering in product, process, or equipment. [researcher inference · strong] Deep-tech deficit · Manufacturing strategy Short origin quote: “I don’t think climate tech has close to lived up to its potential, at least in India” Source note
• Mechanism and reasoning — Owned engineering can support pricing power and export relevance; qualification, yield, reliability, and supplier integration turn invention into repeatable production. Capital creates option value only when it advances one of those observable states. [researcher inference · strong] Advanced-manufacturing report
• Historical and vehicle context — Software-heavy talent and venture portfolios left physical-engineering translation thin; Accel’s later $650 million vehicle and Atoms for X widened capacity, but decision rights, reserves, technical experts, and return targets remain undisclosed. [contemporaneous record · moderate] Unicorn-forecast note
• When to use — Use for robotics, aerospace, defence, industrial AI, energy components, and precision manufacturing with an identifiable buyer and qualification path; do not infer investability from geopolitics, subsidies, demand forecasts, or “deep tech” labels alone. [researcher inference · strong] Robotics note
• How to apply it — Score IP ownership and freedom to operate, technical team depth, TRL, prototype evidence, buyer specification, certification, yield, cycle time, bill of materials, working capital, service liability, export comparison, and the cash required to each gate; stop when proof stalls. [researcher inference · strong]
• Examples —
  • Researcher-applied illustration — Nabhdrishti Aerospace, Scimplify, and Fabheads: Mint maps them to product-, process-, and equipment-led archetypes; it does not show Prakash sourcing, IC sponsorship, technical scores, terms, or outcomes. Mint manufacturing section
  • Researcher-applied illustration — Infra.Market: technology-enabled aggregation and owned plants demonstrate industrial scale, but the public account supports a coordination/full-stack thesis more than proprietary qualification-led manufacturing. Interview note
  • Researcher-applied illustration — Climate technology: Prakash says the category had not come close to its Indian potential; the source does not identify the failed-company denominator or cash losses. Previous-misses note
• Evidence — Four-plus placements include the 2023 deep-tech diagnosis, 2024 robotics interview, 2025 Mint admission and taxonomy, Moneycontrol’s panel account, and Accel’s report. These are largely prospective and institutionally related, not independent realized-return evidence. 2024 robotics source · 2025 panel source
• Limits, failure modes, and exception ledger — IP can be weak, encumbered, easy to design around, or irrelevant to procurement; qualification can outlast reserves; unicorn value is not cash return. Attempted routes included the full report, Mint and Moneycontrol interviews, Business Standard, company cards, offer documents, and current portfolio pages; no IC memos, technical benchmarks, qualification files, rejected-company funnel, vehicle cash flows, or mature outcome cohort were found. Below target for demonstrated application.
• Evolution and contradictions — Broad climate and “Make in India” enthusiasm becomes a narrower hardware/components and proprietary-engineering screen. That is a plausible learning response, but no source proves the earlier misses caused the new process or that qualification governance changed. [researcher inference · strong]
• Modern VC translation (as of 2026-08-20) — What transfers — require owned product, process, or equipment advantage plus a buyer qualification path. What fails under changed 2026 conditions — subsidies, geopolitical demand, or a unicorn forecast cannot substitute for freedom to operate, yield, reliability, procurement, and financing through repeat production. Observable applicability signals — TRL, pilot buyer, certification, yield, cycle time, bill of materials, service liability, and cash to qualification. Most likely misuse — equating national strategy with company defensibility. Analyst-created hypothetical — fund a rare-earth-free motor only after independent IP review and a customer-backed qualification schedule. [researcher inference · strong] 2025 manufacturing screen · Prior-miss boundary

Application-first locally grounded AI

• Rule — Prefer AI that solves a paid, context-specific workflow using local language, behavior, and distribution; fund foundational or sovereign layers only when strategic need, differentiated data rights, technical proof, and capital fit are explicit. [researcher inference · moderate] Talk synthesis, framework 8
• Evidence status — Prakash directly emphasizes applications, voice-first access, Indian-language context, and selective sovereign capacity. The combined hierarchy and any investment-performance claim remain [researcher inference · provisional] because recordings, decision memos, company outcomes, and a rejected-deal denominator are incomplete. AI voice note
• Provenance and origin — The accessible sequence runs from Karya’s data-and-livelihood case in 2023, through 2024–2025 remarks on applications and voice access, to 2026 Bharat1AI and sovereignty discussions. [researcher inference · moderate] Karya note · Bharat1AI note Short origin quote: “We cannot yet build AI business for Bharat with the existing VC model.” Source note
• Mechanism and reasoning — Local language and workflow context improve reach and task fit; application revenue can arrive before frontier-model scale. Sovereign infrastructure may reduce strategic dependency, but consumes capital and requires data, compute, talent, evaluation, and governance that an application does not. [researcher inference · strong]
• Historical and vehicle context — India’s mobile distribution, linguistic diversity, public digital infrastructure, and cost sensitivity create application opportunities, while national AI missions introduce non-venture objectives. Firm, family, grant, and policy roles must remain separate. [researcher inference · strong] India AI mission note
• When to use — Use for workflow AI where language, domain behavior, or distribution materially changes value; consider foundational layers only with a named strategic dependency and credible technical path; do not call every localized interface a moat or every sovereign project a venture return opportunity. [researcher inference · strong]
• How to apply it — Define user job, language/context failure, paid buyer, baseline completion, evaluation set, data rights and compensation, inference-adjusted margin, distribution, safety, and dependency risk; for foundation layers add compute, benchmark, talent, governance, and financing gates. [researcher inference · strong]
• Examples —
  • Researcher-applied illustration — Karya: Prakash discusses data work and livelihoods as a locally grounded AI input; the record does not establish a Prakash investment, vehicle, ownership, or venture return. Karya note
  • Researcher-applied illustration — Bharat1AI: inclusive-economy framing supports local-language and public-purpose reasoning, but not an investable company thesis or independent outcome. Bharat1AI note
  • Researcher-applied illustration — Voice-first startups: Prakash identifies voice as an access path beyond text interfaces; no source supplies a named Prakash-led deal, cohort conversion, retention, or economics. Voice-first note
• Evidence — Four-plus placements span Karya, FCTC remarks, India AI Mission coverage, the voice-first interview, Bharat1AI, and a SuperReturn organizer summary. Several are ecosystem or policy records rather than investment decisions, and some lack full recordings. FCTC note · SuperReturn note
• Limits, failure modes, and exception ledger — Local language may be a feature rather than a moat; data work can externalize labor harms; application margins can collapse under compute or platform dependence; sovereignty can become subsidy-led duplication. Attempted routes covered available captions, organizer pages, Moneycontrol/IndiaAI reporting, Karya and Bharat1AI talks, and firm/profile searches; no full decision memos, deal roster, benchmark pack, application cohort, data-contract audit, or realized outcomes were found. Below target for investment application.
• Evolution and contradictions — The record broadens from inclusive data production to application-first access and selective sovereign capacity. Application efficiency and national infrastructure are not fully consistent capital objectives, so they should be evaluated through separate vehicles and scorecards. [researcher inference · strong]
• Modern VC translation (as of 2026-08-20) — What transfers — start with a paid local workflow, language/context failure, distribution, data rights, and task evaluation. What fails under changed 2026 conditions — localization is not a moat when switching is easy, inference erases margin, data labor is unprotected, or sovereign infrastructure serves a non-venture objective. Observable applicability signals — subgroup task accuracy, paid repeat, licensed data, compensated contributors, inference-adjusted margin, and model-switch resilience. Most likely misuse — conflating inclusive or sovereign policy value with venture return. Analyst-created hypothetical — pilot a voice compliance workflow in one language and separate any public-language infrastructure into another scorecard and vehicle. [researcher inference · strong] AI-for-Bharat constraint · 2026 access limit

Real liquidity and public-market discipline

• Rule — Treat an exit as cash or freely realizable securities, not a valuation headline; approach public markets only with durable core economics, coherent acquisitions, controlled promises, and governance that can withstand recurring disclosure. [researcher inference · strong] IPO-profitability note
• Evidence status — Prakash repeatedly describes IPOs, profitability, and public-market scrutiny as ecosystem disciplines. Model efficacy is [researcher inference · provisional]. Cash proceeds are documented for BlueStone only in bounded offer terms; most portfolio returns, lockups, distributions, and post-listing value are unavailable. Talk synthesis, framework 9
• Provenance and origin — Earlier ecosystem accounts emphasized seed formation and follow-on capital; by 2022–2026, the corpus makes profitability, IPO readiness, domestic liquidity, and fewer funding-dependent companies explicit. [researcher inference · strong] Profitability note Short origin quote: “We should not get ahead of ourselves; we need 5-10 good IPOs.” Source note
• Mechanism and reasoning — Realizations recycle capital, talent, and credibility; public disclosure can force operating and governance discipline. A valuation or listing without durable cash generation, tradable proceeds, or honest acquisition accounting does not establish investor return. [researcher inference · strong]
• Historical and vehicle context — India’s earlier exit scarcity constrained the venture flywheel; deeper domestic markets and founder acceptance later improved routes to liquidity. Realization still depends on security, ownership, OFS allocation, lockups, tax, and fund distribution. [researcher inference · strong] Bengaluru–Mumbai note
• When to use — Use in late-stage follow-ons, acquisition strategy, IPO preparation, and retrospective performance; do not count unicorn status, gross secondary value, M&A announcement, or board association as realized personal or fund return. [researcher inference · strong] Economics boundary
• How to apply it — Reconcile security and vehicle ownership, invested cash, dilution, primary versus secondary shares, OFS proceeds, lockups, tax, distributions, post-listing liquidity, and opportunity cost; separately test profitability, related-party controls, acquisition integration, and disclosure quality. [researcher inference · strong]
• Examples —
  • Documented application — BlueStone: the IPO and Accel-related OFS supply an identifiable liquidity route; allocation to particular funds or Prakash, tax, distributions, and final multiple remain incomplete. IPO reflection · Economics table
  • Researcher-applied illustration — UrbanTouch: a tribunal record reproduces invested and sale-consideration amounts yielding a bounded roughly 0.596x observed ratio before unknown cash flows. Decision lab
  • Researcher-applied illustration — Good Glamm: a content-to-commerce roll-up later entered lender-driven CIRP, illustrating why acquisition scale and funding are not liquidity or public-readiness evidence. Decision lab
• Evidence — Four-plus placements include Prakash’s 2022 and 2026 profitability remarks, 2024 and 2025 IPO forecasts, BlueStone offer/reflection records, UrbanTouch’s tribunal record, and Good Glamm’s NCLT record. Forecasts are dated for later scoring rather than treated as outcomes. IPO forecast note · Robotics/IPO note
• Limits, failure modes, and exception ledger — IPOs can transfer risk rather than create durable value; proceeds can belong to a different vehicle or seller; public prices can fall after lockup; profitability can be engineered by underinvestment. The complete fund cash-flow ledger and Prakash allocation remain unavailable. [researcher inference · strong]
• Evolution and contradictions — The narrative advances from ecosystem capital formation toward real exit and profitability discipline. Yet Prakash also makes time-bounded unicorn and IPO-count forecasts; valuation and count metrics are weaker than the cash-and-quality rule. [researcher inference · strong] Manufacturing forecast
• Modern VC translation (as of 2026-08-20) — What transfers — reconcile invested cash, security, ownership, dilution, primary and secondary shares, lockups, tax, and distributions. What fails under changed 2026 conditions — a valuation, IPO count, gross OFS, or current market capitalization is not a realizable Prakash or fund return without seller and vehicle allocation. Observable applicability signals — filed security terms, named seller, proceeds, lockup expiry, freely tradable value, and fund distribution. Most likely misuse — reporting paper value or aggregate OFS as personal cash. Analyst-created hypothetical — maintain a security-by-vehicle realization ledger through lockup before calling an IPO a win. [researcher inference · strong] 2026 profitability context · Vehicle/economics controls

Ecosystem density as an investment input

• Rule — Treat a location’s repeated interaction among anchor employers, R&D, founders, customers, capital, universities, public infrastructure, and exits as a measurable sourcing and scaling input—not as proof that any city or company will win. [researcher inference · strong] Bengaluru foreword
• Evidence status — Prakash repeatedly states Bengaluru/India ecosystem mechanisms across independent talks, reports, and essays. The model is stronger as a sourcing map than as a causal return predictor because city rankings, counterfactual regions, and company-level attribution are incomplete, so investment application is [researcher inference · insufficient public record] and below target. Talk synthesis
• Provenance and origin — The 2020 Golden Age essay described India’s founder and capital flywheel; later Unboxing Bengaluru work decomposed the city into historical institutions, talent, networks, and global links rather than a single startup-era cause. [researcher inference · strong] Golden Age · Unboxing note Short origin quote: “the network effect in Bengaluru is not just far more apparent but also much more amplified in length and breadth.” Source note
• Mechanism and reasoning — Dense networks lower search, hiring, reference, customer, and knowledge-transfer costs; exits recycle operators and capital; public and research institutions expand the opportunity set. Congestion, groupthink, wage inflation, and policy capture can reverse those benefits. [researcher inference · strong]
• Historical and vehicle context — Bengaluru’s engineering, public-sector, IT-services, multinational, and startup layers predate Accel India; Prakash is a participant, investor, and civic convenor, not the sole author of the ecosystem. Separate these roles and vehicles in every claim. [researcher inference · strong] People map
• When to use — Use for sourcing allocation, talent diligence, regional customer access, and ecosystem-building; do not reject companies outside a celebrated cluster or infer quality from a prestigious address, network, or accelerator. [researcher inference · strong]
• How to apply it — Map anchor institutions, founder recirculation, specialist talent, customer proximity, follow-on capital, research translation, policy reliability, exit recycling, cost, and inclusion; compare the company’s actual dependencies with at least one alternative geography. [researcher inference · strong]
• Examples —
  • Researcher-applied illustration — Bengaluru technology stack: Prakash’s foreword and Unboxing work connect earlier public/R&D institutions, IT services, global capability centers, and startups; the causal weights are not estimated. Foreword note · BBLF note
  • Researcher-applied illustration — India startup flywheel: the Golden Age essay links talent, founder ambition, capital, and exits; it is a joint Accel narrative with survivor and institutional-interest bias. Golden Age note
  • Researcher-applied illustration — Advanced manufacturing: robotics, aerospace, and precision production require laboratories, suppliers, qualification customers, and industrial talent beyond generic startup density. Robotics note
• Evidence — Four-plus placements span the Golden Age essay, Bengaluru report foreword, Unboxing work, Khazanah and IGF talks, and the manufacturing/deep-tech critique. Multiple records are linked to Prakash-led civic organizations, so institutional relatedness is disclosed. Khazanah note · IGF note
• Limits, failure modes, and exception ledger — Ecosystem narratives invite hometown bias, post hoc causality, double counting, and exclusion of weak outcomes. City-level growth does not establish Prakash’s deal returns; the corpus lacks matched regional cohorts, causal estimates, cost-adjusted founder outcomes, and complete failure migration data. [researcher inference · strong] Attempted routes: Golden Age, the Unboxing Bengaluru book excerpts and foreword, BBLF/IGF/Khazanah talks, Bengaluru report material, deep-tech/manufacturing interviews, company ledgers, and regional-result searches. Results: civic mechanisms and macro narratives, but no Prakash sourcing-allocation log, company-level location counterfactual, matched regional cohort, cost-adjusted causal estimate, or attributable return; investment application is below target. Unboxing note · Talk coverage · Investment searched denominator
• Evolution and contradictions — The thesis evolves from India-scale optimism to a richer, historically layered Bengaluru account and then to sector-specific infrastructure gaps. The correction is to measure the exact density a company needs rather than “Bengaluru” as one undifferentiated advantage. [researcher inference · strong]
• Modern VC translation (as of 2026-08-20) — What transfers — map the specific talent, customer, laboratory, supplier, capital, policy, and exit dependencies a company needs. What fails under changed 2026 conditions — city prestige is a weak proxy when remote work, cost, or sector-specific infrastructure changes which dependencies must be local. Observable applicability signals — time to hire, expert referrals, customer access, lab or supplier proximity, follow-on availability, cost, and operator recirculation. Most likely misuse — turning civic advocacy or a ranking into a picking signal. Analyst-created hypothetical — compare an aerospace supplier’s qualification network in two regions instead of assigning a Bengaluru premium. [researcher inference · moderate] Bengaluru network evidence · Deep-tech infrastructure gap

Venture methods with vehicle separation

• Rule — Apply venture habits—problem selection, operator sourcing, time-boxed experiments, milestones, and rapid feedback—to public or social problems only after specifying mandate, capital type, decision rights, beneficiary, conflicts, governance, and a non-financial scorecard. [researcher inference · strong] Talk synthesis, framework 10
• Evidence status — Prakash’s ACT Grants, policy, AI-inclusion, and ecosystem work documents transfer of venture methods. The separation rule is a researcher safeguard derived from his overlapping roles and is [researcher inference · provisional]; public records do not supply complete grant ledgers, recusals, outcome evaluations, or Prakash-only decisions. ACT Grants note
• Provenance and origin — The corpus moves from investment-platform building into pandemic grants, Karnataka and national startup policy, workforce/data inclusion, and city institutions. Each adds public purpose without converting grants or civic results into venture returns. [researcher inference · strong] FCTC note Short origin quote: “We focused more on solutions that were ready to go to market and had great teams” Source note
• Mechanism and reasoning — Venture methods can accelerate search, recruit operators, stage scarce capital, and expose failure early. Public and social systems have plural beneficiaries, due process, distributional effects, and accountability that market prices cannot resolve, so separate authority and metrics are essential. [researcher inference · strong]
• Historical and vehicle context — Prakash has acted through Accel, ACT Grants, family/philanthropic capital, civic organizations, and policy forums. A title or association does not prove investment, donation, board action, procurement influence, or personal economic benefit. [researcher inference · strong] Attribution rules
• When to use — Use for emergency response, ecosystem infrastructure, workforce/data inclusion, and policy pilots with a measurable public problem; do not use venture speed to bypass procurement, legal review, privacy, community consent, democratic accountability, or conflicts. [researcher inference · strong]
• How to apply it — Publish the problem, legal authority, funder and instrument, selector, recusal rules, beneficiary, milestone, data rights, harm metric, independent reviewer, handoff or sunset, and results; keep venture sourcing and portfolio economics in a separate ledger. [researcher inference · strong]
• Examples —
  • Documented application — ACT Grants: the pandemic response used founder networks and rapid grants; speed is plausible process evidence, while grant attribution, full denominator, and causal health outcomes remain incomplete. ACT Grants note
  • Researcher-applied illustration — Karnataka/startup-policy work: Prakash discusses founders and government working together; the evidence supports advocacy and participation, not unilateral authority or private investment return. FCTC note
  • Researcher-applied illustration — Karya/Bharat1AI: data-work and inclusive-AI cases show beneficiary and sovereignty objectives that require labor, privacy, and public-value measures separate from venture valuation. Karya note · Bharat1AI note
• Evidence — Four-plus placements span ACT Grants, FCTC, Karya, Bharat1AI, MyGov and Sahyadri public forums, and Budget 2024 writing. They support role breadth and methods, but most are participant accounts or event records without independent impact evaluation. MyGov note · Budget note
• Limits, failure modes, and exception ledger — Speed can weaken due process; network sourcing can exclude outsiders; private investors may shape public priorities; success metrics can privilege visibility over welfare. Attempted routes covered talks, captions, policy/event pages, ACT material, company and nonprofit records, and role disclosures; no complete grant/beneficiary ledger, procurement files, recusal register, counterfactual evaluation, or unified conflict statement was found. Below target for independently evaluated outcomes.
• Evolution and contradictions — Hands-on investing habits expand into crisis, policy, and ecosystem work, while vehicle separation becomes more important as roles multiply. No source shows a formal integrated conflict architecture; the separation and publication requirements are analytic safeguards, not claimed Prakash procedure. [researcher inference · strong]
• Modern VC translation (as of 2026-08-20) — What transfers — use problem selection, operator sourcing, staged experiments, milestones, and rapid feedback. What fails under changed 2026 conditions — public or social work cannot inherit venture speed when due process, plural beneficiaries, procurement, privacy, or democratic accountability require different authority and metrics. Observable applicability signals — published mandate, funder, instrument, selector, recusal, beneficiary, harm metric, independent review, and sunset. Most likely misuse — converting public influence into proprietary deal flow or venture credit. Analyst-created hypothetical — run a time-boxed public-health grant with an independent selector and a separate non-financial outcome ledger. [researcher inference · strong] ACT practice · Policy-role boundary

Evidence gaps and candidate models

• Inventory reconciliation — Fifteen supportable models survive synonym merging across the complete 100-note denominator: 74 deduplicated talk families and 26 priority written records, reconciled against the deal/opportunity corpus. “Digital first” is merged into full stack and omnichannel; “founder resilience” into founder learning; “venture debt” into capital efficiency; “India as a global lab” into customer, manufacturing, AI, and ecosystem models; “philanthropic venture” into vehicle-separated public methods. Talk denominator · Writing denominator · Investment denominator
• Signal-to-model reconciliation map —
  • Theme and sourcing signals from the prepared-mind, local-market, and ecosystem corpus map to Prepared mind, Reachable customer, and Ecosystem density. Talk frameworks 1–2 · Writing synthesis
  • Founder, intervention, and governance signals map to Founder learning, Permissioned mentoring, and Vehicle separation. Venture-lifecycle boundary · Attribution controls
  • Operating and scaling signals map to Evidence and capital efficiency, Fragmentation, Minimum full stack, and Density-gated omnichannel. Historical operating rules · Decision labs
  • Duration and vehicle signals map to Patient capital and Fund-scale fit. Vehicle evidence · Talk framework 4
  • Frontier and realization signals map to IP-led manufacturing, Locally grounded AI, and Real liquidity. Talk frameworks 7–9 · Portfolio forensics
• Philosophy-index reconciliation — The direct Mental models and maxims index covers all 15 stable headings, while the philosophy also uses lifecycle sections: Snapshot/edge thesis links models 1, 11, 12, and 14; Core philosophy links models 1–15 at belief level; Sourcing links models 1, 2, 3, and 14; Picking links models 3–7, 11, and 12; Deal and portfolio mechanics links models 4 and 7–13; Vehicle and incentive fit links models 8, 10, 13, and 15. This is a crosswalk, not evidence that the philosophy document independently validates each model.
• Global evidence gaps — No complete Prakash-authored investment ledger, rejected-deal denominator, IC memo set, partner voting record, board-minute corpus, ownership history, reserve policy, fund cash-flow statement, attribution schedule, realized-return denominator, conflict/recusal register, or matched failure cohort was located. These omissions block claims of personal track record, model-level alpha, or causal operating value-add. Investment uncertainty ledger · Talk evidence gaps · Writing unresolved record
• Source and retrieval gaps — Several videos lack primary recordings or complete transcripts; event pages and ASR can truncate caveats; archived Erasmic pages are contemporaneous but promotional; firm essays are jointly authored and retrospective; private-company filings and news do not reveal full capital or governance. Each below-target model records its inspected routes and missing artifacts in its own exception ledger. Talk retrieval hierarchy · Writing evidence boundary
• Candidate: write the falsifier beside every forecast — The corpus supports dated forecasts and later contradiction checks, but not a repeated Prakash-stated forecasting protocol distinct from prepared-mind, manufacturing, AI, and liquidity models. Keep as a candidate until a primary source shows a systematic pre-mortem, threshold, review date, and action. Historical operating rules · Talk contradictions
• Candidate: collective credit over individual hero attribution — Prakash often speaks institutionally and credits founders or ecosystems, but this is presently an attribution discipline rather than a sufficiently evidenced investment mechanism with three decision cases and an adverse boundary. Attribution controls · Joint-authorship boundary
• Candidate: premium-to-mass cross-subsidy — A small subset of inclusive-business material suggests higher-value demand may finance broad access, but the corpus lacks three Prakash-linked decisions, unit economics, beneficiary outcomes, and a failure control. It remains below inventory threshold. Biopeak note · AI-for-Bharat financing boundary
• Candidate: governance independence as a standalone model — Board independence, longer diligence, and role boundaries are important, but public cases do not show three Prakash-specific interventions and outcomes. Governance is embedded in mentoring, patient capital, liquidity, and vehicle separation rather than inflated into a sixteenth model. Diligence change · Governance boundary

Direct writings and first-person institutional record

• ★★★ | Erasmic — Funding Philosophy (original: http://www.erasmic.com/about_philosophy.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Fabout_philosophy.html&output=json | captured 2008-08-20 01:45:54 UTC | contemporaneous unbylined firm page recovered from Common Crawl ARC; reliability A for Erasmic's position, access full; exact ARC metadata retained in note | Primary institutional doctrine on seed-stage capital and support; team/vehicle voice, not Prakash-personal authorship.
• ★★★ | Erasmic — Investment Strategy (original: http://www.erasmic.com/investment_strategy.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Finvestment_strategy.html&output=json | captured 2008-07-20 00:49:03 UTC | contemporaneous unbylined firm page recovered from Common Crawl ARC; reliability A for Erasmic's position, access full; exact ARC metadata retained in note | Earliest recoverable strategy screen; preserve company stage, geography, and institutional-team attribution.
• ★★★ | Erasmic — Our Value Add (original: http://www.erasmic.com/about_value_add.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Fabout_value_add.html&output=json | captured 2008-08-20 01:45:20 UTC | contemporaneous unbylined firm page recovered from Common Crawl ARC; reliability A for Erasmic's position, access full; exact ARC metadata retained in note | Specifies claimed operating support and governance contribution; advocacy must be checked against company and board evidence.
• ★★★ | Erasmic — Overview (original: http://www.erasmic.com/about_overview.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Fabout_overview.html&output=json | captured 2008-08-20 01:46:30 UTC | contemporaneous unbylined firm page recovered from Common Crawl ARC; reliability A for firm plan, access full; exact ARC metadata retained in note | Proposed $10 million fund, four-to-six-company focus, $100,000–$500,000 checks, and senior-management time; plan is not closing/performance evidence.
• ★★★ | Erasmic — Investment Area (original: http://www.erasmic.com/investment_area.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Finvestment_area.html&output=json | captured 2008-08-20 01:47:03 UTC | contemporaneous unbylined firm page recovered from Common Crawl ARC; reliability A for Erasmic's position, access full; exact ARC metadata retained in note | Primary market/technology scope and exclusion evidence for reconstructing the early funnel.
• ★★★ | Erasmic — Track Record (original: http://www.erasmic.com/about_track_record.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Fabout_track_record.html&output=json | captured 2008-08-20 01:48:08 UTC | contemporaneous firm intervention ledger recovered from Common Crawl ARC; reliability A for what Erasmic claimed, access full; exact ARC metadata retained in note | Names company assistance and exits including Mu Sigma/PRO; valuable team-behavior record but not audited attribution or return evidence.
• ★★ | Erasmic — Positioning (original: http://www.erasmic.com/about_positioning.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Fabout_positioning.html&output=json | captured 2008-07-20 00:48:39 UTC | contemporaneous unbylined firm page recovered from Common Crawl ARC; reliability A for Erasmic's position, access full; exact ARC metadata retained in note | Differentiation and ecosystem posture; useful control against later Accel retrospective branding.
• ★★ | Erasmic — Indian Landscape (original: http://www.erasmic.com/investment_indian_landscape.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Finvestment_indian_landscape.html&output=json | captured 2008-08-20 01:47:37 UTC | contemporaneous unbylined firm page recovered from Common Crawl ARC; reliability A for the dated thesis, access full; exact ARC metadata retained in note | Ex-ante India opportunity framing before later outcomes; forecasts require period-matched base-rate controls.
• ★★ | Erasmic — Founding Team (original: http://www.erasmic.com/team_founding.html) | https://index.commoncrawl.org/CC-MAIN-2008-2009-index?url=erasmic.com%2Fteam_founding.html&output=json | captured 2008-07-20 00:48:55 UTC | contemporaneous institutional team/biography page recovered from Common Crawl ARC; reliability A for firm representation, access full; exact ARC metadata retained in note | Primary team-composition and role boundary; biography evidence, not a Prakash-authored investment rule.
• ★★★ | Prashanth Prakash — Accel profile and live relationships | https://www.accel.com/team/prashanth-prakash | current page accessed 2026-08-19 | first-person profile plus official mutable affiliation graph; reliability A for Prakash's statements and Accel's current labels; access full; locators: About / Get to know / Relationships | Direct voice on early product work, Erasmic, focus, learning process, and credit; embeds 28 current company links and three publications but is not a historical deal ledger.
• ★★★ | Celebrating BlueStone’s IPO: A decade of design, trust, and quiet endurance | https://www.accel.com/news/celebrating-bluestones-ipo-a-decade-of-design-trust-and-quiet-endurance | 2025-08-18 | sole-authored Accel portfolio essay; reliability A for Prakash's dated account, B for retrospective deal narrative; access full; locators: Why we leaned in early / hyperlocal thesis / closing signature | Highest-value Prakash-specific retrospective reconstruction of the 2011 entry thesis and IPO reflection; control against the prospectus, ownership, team, and board records.
• ★★★ | Building a digital CPG giant in India: Our investment in MyGlamm | https://www.accel.com/news/building-a-digital-cpg-giant-in-india-our-investment-in-myglamm | 2021-08-04 | co-authored Accel investment essay; reliability A for the team statement, B for retrospective company account; access full; locators: byline and paras. 15–30 | Written evidence on D2C category mapping, founder assessment, growth-stage rationale, COVID pivot, and shared Anand Daniel/Prakash/Rachit Parekh attribution.
• ★★★ | Transforming the Seafood Supply Chain: Our Investment in Captain Fresh | https://www.accel.com/news/transforming-the-seafood-supply-chain-our-investment-in-captain-fresh | 2021-07-20 | co-authored Accel investment essay; reliability A for the team statement, B for retrospective rationale; access full; locator: byline and thesis body | Written sourcing and B2B supply-chain thesis; keeps Barath Shankar Subramanian, Prakash, and Rachit Parekh credit joint.
• ★★★ | Innovation, sustainability to fuel transformative phase | https://economictimes.indiatimes.com/news/india/innovation-sustainability-to-fuel-transformative-phase-accels-prashanth-prakash/articleshow/106425708.cms | 2023-12-31 | Prakash-authored op-ed; reliability A for his dated views, B/C for unsupported forecasts; access full; locators: byline and body paras. 128–147 | Direct writing on AI, manufacturing-as-a-service, green energy, Global South expansion, and 2024 expectations; later results and changed views are required controls.
• ★★★ | India’s manufacturing journey is set to get better in 2025 | https://www.moneycontrol.com/news/opinion/india-s-manufacturing-journey-is-set-to-get-better-in-2025-12897899.html | 2024-12-30 | Prakash-authored op-ed; reliability A for views, B/C for cited market forecasts without underlying links; access full; locators: byline / four pillars / AI-human collaboration | Direct written evolution toward precision engineering, robotics, R&D-led manufacturing, defence, and hybrid AI-human work.
• ★★★ | The Golden Age of Indian Startups, VCs | https://medium.com/accel-india-insights/the-golden-age-of-indian-startups-vcs-4aebdb78fffb | 2020-01-09 | co-authored Accel India history; reliability A for the authors' account, B for retrospective firm narrative; access full | Direct institutional chronology for Erasmic, the first Accel India fund, internet adoption, Flipkart, and ecosystem development; preserve co-author and 2008/2009 date boundaries.
• ★★★ | Decoding Omnichannel: Strategies for D2C Brands | https://www.seedtoscale.com/decoding-omnichannel-report | 2024-05-09 launch; page undated | six-author Accel/Fireside/Redseer report; reliability B for framework, C for unsourced forecasts/company-selected cases; access substantial rendered page, separate download unresolved | Adds a staged online-to-offline expansion framework, brand/demand signals, and Lenskart/BlueStone/AgroStar cases; no sentence-level Prakash authorship or failed-rollout denominator.
• ★★★ | India’s Advanced Manufacturing Leap: Pioneering the Future of IP-Driven Manufacturing | https://events-in.accel.com/advanced-manufacturing-report | 2025-08-14 | four-author Accel/Praxis research report; reliability B for taxonomy and dated thesis, C for forecasts/promotional comparisons; access full rendered report, separate download registration-gated | Defines product, equipment-led process, and process innovation plus IP/performance screens; controls the 2023–24 manufacturing thesis but lacks company-level decision/economics evidence.
• ★★★ | A museum in Bengaluru will showcase India’s technology, innovation & startup feats | https://timesofindia.indiatimes.com/city/bengaluru/a-museum-in-bengaluru-will-showcase-indias-technology-innovation-startup-feats/articleshow/96632714.cms | 2022-12-31 | Prakash-bylined civic/technology-history op-ed with TNN page credit; footer identifies Prakash as writer; reliability A for his proposal and views, B/C for historical generalizations; access full | Earliest located Prakash-bylined piece; explains the Technology and Innovation Museum proposal and a public-memory/civic-infrastructure thesis that must not be recast as venture doctrine.
• ★★★ | Budget 2023: Focus on emerging sectors praiseworthy but startups need a helping hand | https://www.moneycontrol.com/news/opinion/budget-2023-focus-on-emerging-sectors-praiseworthy-but-startups-need-a-helping-hand-9988711.html | 2023-02-02 | Prakash-authored policy op-ed; reliability A for views, B/C for causal and macro claims; access full | Connects infrastructure, agritech, green energy, AI, angel tax, GIFT IFSC, and capital-gains policy to startup formation; the page identifies his Accel affiliation, making the recommendations interested but not a formal conflict disclosure.
• ★★★ | Bullish on India: Watch out for startups bringing innovation and disruption to manufacturing | https://www.moneycontrol.com/news/opinion/bullish-on-india-watch-out-for-startups-bringing-innovation-and-disruption-to-manufacturing-11328981.html | 2023-09-07 | Prakash-authored manufacturing op-ed; reliability A for dated views, B/C for market forecasts; access full | Intermediate statement in the 2023–25 manufacturing thesis, useful for tracing policy, infrastructure, supply-chain, founder, and venture-capture assumptions.
• ★★★ | Seizing global opportunities: Budget 2024’s vision for India’s future | https://www.moneycontrol.com/news/opinion/seizing-global-opportunities-budget-2024s-vision-for-indias-future-12777184.html | 2024-07-24 | Prakash-authored policy op-ed; reliability A for views, B/C for causal and forecast claims; access full | Updates the public-policy thesis across manufacturing, climate, agriculture, digital infrastructure, skills, and global competitiveness; policy enthusiasm is not portfolio performance.
• ★★★ | Bengaluru Innovation Report 2024 — signed Prashanth Prakash foreword | https://cdn.tracxn.com/marketing-campaigns/Bangalore_Innovation_Report_-_2024_A4AAy88hUDJthl4jJ69_N.pdf | 2024 | signed foreword on PDF p. 5; report body has separate institutional/data authorship; reliability A for foreword, B/C for report claims; access full PDF | Direct civic/ecosystem statement on startup, GCC, talent, government, and ten-year city planning; do not attribute the 66-page report body to Prakash.
• ★★★ | Unboxing Bengaluru: The City of New Beginnings | https://www.penguin.co.in/book/unboxing-bengaluru/ | 2023-11-30 ebook/original hardback; UnboxingBLR reports 2022 and paperback reissued January 2026 | co-authored book; reliability A for publisher metadata and authorized excerpts, B for catalog metadata; access partial via authorized Penguin/Moneycontrol/TOI excerpts; locators in note | Analyzable excerpts cover Bengaluru's network density, pay-it-forward culture, founder life, and urban fault lines; the unopened body and Goyal/Prakash sentence-level authorship remain unresolved.
• ★★ | India’s next leap in global relevance will come from what we build, not just the software we write | https://www.linkedin.com/posts/prakashprashanth_accel-advancedmanufacturing-defencetech-activity-7351626579396976643-vKyk | created 2025-07-17; current body accessed 2026-08-19 | direct LinkedIn post; reliability A for Prakash's current wording, provisional for exact creation-date wording because LinkedIn marks it Edited and exposes no revision history; access full current body; date decoded from activity ID | Manufacturing thesis associated with a pre-report post date; whether the current exact wording preceded Accel's August report remains unresolved and is not performance evidence.
• ★★ | National Startup Day reflection — India’s real advantage will come from depth | https://www.linkedin.com/posts/prakashprashanth_press-trust-of-india-activity-7417937321452257280-qaxw | 2026-01-16 | direct LinkedIn post linking a PTI interview; reliability A for Prakash's post, B for PTI-reported quotations; access full public post, linked PTI body available through authorized syndication; date decoded from activity ID | Latest located direct writing on IP-led outcomes, applied science, manufacturing, and ecosystem depth; separate the authored post from the reported interview it endorses.

Interviews, talks, video, and evolving direct voice

• ★★★ | The Accel effect | https://www.fortuneindia.com/venture/the-accel-effect/101864 | Published/Event: 2018-07-13 online; originally published in the 2018-03-15 to 2018-06-14 special issue | Format: reported firm profile with multi-speaker interview excerpts; reliability: A for named quotations, B for narrative, Tracxn estimates, company figures, and performance framing; access: full rendered article; underlying interviews and datasets unavailable | Unique value: Fortune's central skeptical control is liquidity: it says much apparent success was notional, reports more than 35 exits with 18 M&As and estimated 1x–12x returns, and notes that Myntra and TaxiForSure consideration was buyer stock rather than cash. The figures are attributed to Tracxn estimates, not audited vehicle statements. | Deep note.
• ★★★ | "Early Stage Focus Will Remain, But More Money At Our Disposal" | https://www.vccircle.com/early-stage-focus-will-remain-more-money-our-disposal | Published/Event: 2012-04-30 on the current page; the interview concerns the 2008 Erasmic–Accel combination and appears to have been migrated or redated | Format: edited email interview / Q&A; reliability: A for Prashanth Prakash's published answers, B for Shrija Agrawal's framing; access: full rendered Q&A; no original email chain or edit history | Unique value: Prakash presented the Erasmic–Accel combination as an institutional-capability decision: Accel's domain expertise, global founder appeal, and cross-hub relationships would supplement a locally led India team. These are expected mechanisms, not measured company or fund outcomes. | Deep note.
• ★★★ | Accel aims to invest in 15–25 Indian startups in 2024 | https://www.business-standard.com/companies/start-ups/venture-capital-major-accel-aims-to-invest-in-25-indian-startups-prashanth-124032500406_1.html | Published/Event: 2024-03-25 | Format: reported interview; reliability: A for Prakash-attributed quotations; B for reporter framing and third-party Tracxn data; access: full rendered text | Unique value: Prakash said Accel still had substantial dry powder, expected 15–25 Indian investments in 2024, and generally launched a fund every three to four years after deploying the current vehicle. This is a plan stated during Fund VII, not evidence of actual 2024 investments or a commitment to a launch date. | Deep note.
• ★★★ | ACT Grants helped create disruptive, scalable solutions to fight Covid-19 | https://indiacsr.in/act-grants-helped-create-disruptive-scalable-solutions-to-fight-against-covid19-prashanth-prakash/ | Published/Event: 2020-08-14 | Format: edited interview with Apresh Mishra; reliability: A for Prakash-attributed answers; B for India CSR framing and ACT self-report; access: full body indexed in search; direct page returned HTTP 429 during final open | Unique value: Prakash says ACT emerged after more than 1,000 startup participants gathered in a Telegram group and 15–20 venture, private-equity, and investment-banking groups created a structured grant mechanism. The claimed ₹100 crore corpus came from venture participants, industrialists, and entrepreneurs; no audited deployment ledger is supplied here. | Deep note.
• ★★ | Over 250 Indian AI startups and a voice-first opportunity | https://www.moneycontrol.com/news/business/startup/over-250-indian-ai-startups-emerged-during-last-18-months-says-accel-s-prashanth-prakash-13101983.html | Published/Event: 2025-06-06; event: 2025-06-04 | Format: reported joint fireside remarks at Accel AI Summit; reliability: A for the one Prakash quotation; B for attributed paraphrase; C for the unattributed startup count; access: full article, not full session transcript | Unique value: Moneycontrol reports Prakash's claim that more than 250 Indian AI startups emerged in 18 months. The article supplies no dataset, definition, cutoff, deduplication, or source for the count, so it is retained as a reported claim rather than a verified denominator. | Deep note.
• ★★★ | Unboxing Bengaluru: Memory, Metropolis, and the Making of a New City — BBLF2025 | https://www.youtube.com/watch?v=ik8XmpilX7Y | Published/Event: 2025-10-06 | Format: 49:20 Book Brahma Literature Festival panel with Prashanth Prakash, Kannada translator Prathiba Nandakumar, and moderator Suresh Narasimha; reliability: A for direct speaker substance, B for exact wording in the third-party auto-generated caption mirror, and insufficient for unsourced market shares, forecasts, or third-party performance; access: partial; public auto-captions through 49:12 and YouTube's title/date/runtime metadata were inspected, but creator-caption provenance was unavailable and audio was not independently checked | Unique value: Prakash says the book was meant to help both established and newly arrived residents understand Bengaluru, build civic identification, and confront neglect of cities even though cities are national economic engines. | Deep note.
• ★★★ | A Parallel Universe — BECO Capital panel with Vani Kola and Prashanth Prakash | https://www.youtube.com/watch?v=LoMTGwV2Vts | Published/Event: exact upload date unavailable; the YouTube page rendered “9 years ago” on 2026-08-19 | Format: 39:07 conference panel / joint interview; reliability: A for the speakers' direct statements, B for exact wording in a third-party auto-generated caption mirror, and insufficient for self-reported fund performance, valuations, or deal attribution; access: partial; the complete public auto-caption mirror was inspected, but YouTube returned a 401/sign-in bot challenge, creator-caption provenance was unavailable, and audio was not independently checked | Unique value: Prakash says his first India venture was an early internet company and that he and his business partner both exited companies in the same quarter. They chose a small seed fund over another startup to apply their operating lessons across multiple companies. The account supplies no company, transaction, or vehicle documentation. | Deep note.
• ★★★ | Should founders move from Bengaluru to Mumbai to build? — Prashanth Prakash and Sanjay Nayar | https://www.youtube.com/watch?v=5-jWhGN422c | Published/Event: 2024-08-14 | Format: 20:44 joint Moneycontrol panel with Prashanth Prakash, Sanjay Nayar, and moderators Bodhisatva Ganguli and Chandra Ranganathan; reliability: A for direct speaker statements, B for exact wording in the third-party auto-generated caption mirror, and insufficient for uncorroborated market counts or predictions; access: partial; public auto-captions through 20:38 and YouTube's title/date/runtime metadata were inspected, but creator-caption provenance was unavailable and audio was not independently checked | Unique value: Prakash says the first startup IPO cohort disappointed, but founders and investors quickly learned the public-market discipline: constrain promises, deliver consistently, and combine growth with profits. He uses Zomato as the positive playbook. | Deep note.
• ★★ | Becoming an Inclusive $30 Trillion Economy by 2047 | https://www.youtube.com/watch?v=mVW25EWaa5A | Published/Event: 2026-05-19 | Format: 29:37 four-speaker Bharat1AI fireside panel; reliability: A for official Bharat1AI metadata and participant list; B/C for meaning recovered from a Hindi-script phonetic automatic-caption track; no exact English-speech quote grade; access: full automatic captions through approximately 29:21, but spoken English is rendered phonetically in Devanagari; Prakash's explicit 00:19–04:41 handoff is used for cautious paraphrase only | Unique value: Bharat1AI presents the event as a discussion among Prashanth Prakash, Mohandas Pai, Ashutosh Sharma, and Abhay Singhal about an inclusive $30 trillion Indian economy by 2047. That number is the publisher's session frame, not automatically a Prakash forecast. | Deep note.
• ★★★ | India can be a global lab for scalable longevity solutions | https://www.livemint.com/news/india/india-has-the-potential-to-be-a-global-lab-for-scalable-longevity-solutions-accel-s-prashanth-prakash-11753435048181.html | Published/Event: 2025-07-28 | Format: edited investment interview; reliability: A for Prakash's published answers; B for framing and company descriptions; access: full rendered Q&A | Unique value: Prakash frames the Biopeak decision around the convergence of diagnostics, AI, and longevity science, with a precision-health system adapted to Indian biology, environment, and lifestyle. The answer is a post-investment thesis, not the original memo or clinical validation. | Deep note.
• ★★★ | Pitching Jewelry for the Digital Age: Inside BlueStone’s Original Deck | https://www.seedtoscale.com/podcast/pitching-jewelry-for-the-digital-age-inside-bluestones-original-deck | Published/Event: 2025-09-11 in the canonical RSS feed | Format: 40:29 joint podcast/video plus a publisher-linked copy of the 2011 pitch deck; reliability: A for the deck’s contemporaneous contents; A for direct speech if speaker-verified; B for publisher retrospective summaries; access: partial — official audio, page, and linked deck accessible; no transcript/captions; supplied PDF contains 25 PDF pages numbered 1–24 and 33 rather than the publisher-described complete 33-page deck | Unique value: The 2011 deck’s ex-ante thesis combined high ticket size and margins, a shift toward branded/lightweight/fashion jewelry, a nascent online channel, and an expected low-inventory model. It proposed starting with jewelry and later expanding into watches and lifestyle accessories. | Deep note.
• ★★★ | Epicentre Of Climate-Change Tech Will Move Into India | https://www.moneycontrol.com/news/environment/epicentre-of-climate-change-tech-will-move-into-india-prashanth-prakash-9548031.html | Published/Event: 2022-11-17 on the page header; footer says first published 2022-11-16 | Format: edited interview excerpts / Q&A; reliability: A for Prashanth Prakash's published answers, B for Darlington Jose Hector's framing and cited market figure; access: full rendered article body; an unrelated expandable vaccine-FAQ module interrupts the first answer and is excluded | Unique value: Prakash said ACT grew from a COVID-response coalition into a multi-cause platform after participants saw startup-built solutions as applicable to broader social problems. That is an institutional-origin account, not evidence that the platform caused social outcomes. | Deep note.
• ★★ | Evening Keynote at the CMU India Conclave | https://www.youtube.com/watch?v=CfXEB3aMN5U | Published/Event: event dated 2018-07-14 in the official title; video uploaded 2018-07-24 | Format: 56:03 Carnegie Mellon University event video; reliability: A for CMU provenance and title; B for Deccan Herald event context; no speech-content grade because no transcript or playable media was accessible in this run; access: metadata and independent event report only; direct media/captions and multiple transcript fallbacks failed | Unique value: Carnegie Mellon University's official channel titles the 56:03 item an “Evening Keynote by Prashanth Prakash” at the CMU India Conclave in Bangalore on July 14, 2018. That verifies an appearance record, not the substance of the talk. | Deep note.
• ★★ | Indian Startups: Will 2023 Be Better Than 2022? — CNBC-TV18 Young Turks | https://www.youtube.com/watch?v=60ADJAeZbzU | Published/Event: 2023-01-07 | Format: 26:50 joint television interview with Mohit Bhatnagar and Prashanth Prakash; reliability: A for direct speech if speaker-verified; B for CNBC-TV18 title and description; access: severely partial — title, date, runtime, participants, and topic accessible; no transcript, captions, chapter list, audio, or speaker-attributed report recovered | Unique value: CNBC-TV18 identifies the appearance as a January 7, 2023 Young Turks discussion in which Sequoia India managing director Mohit Bhatnagar and Accel partner Prashanth Prakash shared their outlook for Indian startups in 2023. That establishes the forum and participants, not either investor's answer to the title question. | Deep note.
• ★★★ | Coffee with Ashwath — startups, collective philanthropy, and co-building | https://www.youtube.com/watch?v=HurN0DZzJ7A | Published/Event: 2023-04-03 (YouTube premiere; podcast syndication dates the audio episode 2023-03-31) | Format: 22:41 joint video interview with C. N. Ashwath Narayan; reliability: A for Prashanth Prakash's direct statements, B for exact wording in the third-party auto-generated caption mirror, and insufficient for impact, economic, and priority claims; access: partial; the public mirror's auto-captions cover 22:35 of the 22:41 upload and a syndicated publisher listing was inspected, but creator-caption provenance was unavailable and audio was not independently checked | Unique value: Prakash describes ACT's pandemic model as a coalition of investors/donors, NGOs, and startup founders with technology and operating ideas. His claimed advantage is speed through coordination: the actors could move faster together than independently. | Deep note.
• ★★★ | Will look at global crypto opportunities; there are regulatory challenges in India | https://www.moneycontrol.com/news/business/interview-will-look-at-global-opportunities-for-crypto-there-are-regulatory-challenges-in-india-accel-indias-prashanth-prakash-7731971.html | Published/Event: 2021-11-17 | Format: edited interview at Bengaluru Tech Summit; reliability: A for attributed answers; B for reporter framing; C for forecasts and unsourced talent shares; access: full text | Unique value: At the 2021 funding peak, Prakash saw internet IPOs as a catalytic proof of Indian exit capacity that could attract more capital and support experiments in longer-cycle space and climate technology. This was a prospective claim, not evidence that those new bets later returned capital. | Deep note.
• ★★ | Data Science & AI for Longevity | https://www.youtube.com/watch?v=pEuBOLNAAg0 | Published/Event: 2026-01-07 | Format: 26:47 two-person fireside chat at RISE for Healthy Aging 2025; reliability: B for publisher metadata and joint discussion summary; C for any speaker-level allocation not stated by the publisher; access: partial — complete publisher description and metadata recovered; transcript body and playable media unavailable | Unique value: Longevity India says Prashanth Prakash and AI/ML specialist Ram Prakash H jointly explore data science and AI in longevity and preventive care. The publisher's repeated “they discuss” wording establishes Prakash as a substantive co-discussant, not merely the session host. | Deep note.
• ★★★ | A Toolkit for Decoding Omnichannel — Accel India launch conversation | https://www.youtube.com/watch?v=xT8pUyCc9eM | Published/Event: 2024-06-10 | Format: 27:58 joint panel with Prashanth Prakash, Kanwaljit Singh, and moderator Mangalam Maloo; reliability: A for direct speaker statements, B for exact wording in the third-party auto-generated caption mirror, and insufficient for report forecasts or portfolio outcomes; access: partial; complete public auto-captions through 27:50 and YouTube's title/date/runtime metadata were inspected, but creator-caption provenance was unavailable and audio was not independently checked | Unique value: Prakash distinguishes the launch document from a conventional market report: after a landscape of “trailblazers,” it is meant to help brands decide readiness, choose an offline format, pilot, select metrics, and graduate toward national or multi-format expansion. | Deep note.
• ★★★ | There is no real ecosystem for deep tech in India | https://www.moneycontrol.com/news/business/startup/there-is-no-real-ecosystem-for-deep-tech-in-india-accels-prashanth-prakash-10728441.html | Published/Event: 2023-06-02 | Format: reported event remarks; reliability: A for quoted remarks; B for reporter framing; C for unsourced ecosystem generalizations and third-party figures; access: full article; no event recording/transcript located | Unique value: Prakash’s 2023 deep-tech diagnosis is ecosystem-wide: modest corporate and government R&D, an early financing cliff after $1–3 million, weak university translation incentives, missing industry links, and poor acquisition demand for intellectual property. | Deep note.
• ★★ | Guest of the Week — Prashanth Prakash on Startup Karnataka | https://www.youtube.com/watch?v=PY7YEqvIx_Y | Published/Event: undated — the public watch/search surfaces did not expose an upload date | Format: 55:24 one-to-one Doordarshan Chandana television interview; reliability: A for direct speech if speaker-verified; B for official broadcaster title and description; access: severely partial — exact title, runtime, channel, guest role, and description hashtags accessible; no transcript, captions, chapters, audio, or detailed recap recovered | Unique value: Doordarshan Chandana presents Prashanth Prakash as the sole “Guest of the Week” and identifies him as chairman of the Vision Group on Startup. This confirms a substantive Prakash appearance rather than a passing mention. | Deep note.
• ★★★ | Entrepreneur India Conclave 2018 — sector and selection remarks | https://india.entrepreneur.com/leadership/what-indias-investors-think/336321 | Published/Event: 2018; retrospective article published 2019-06-28 | Format: edited multi-investor event excerpts; reliability: A for attributed Prakash quotations; B for editorial transcription; C for unsourced portfolio counts; access: full excerpt, not full recording/transcript | Unique value: Prakash reduces early-stage selection to three variables: latent/developing market opportunity, entrepreneurial-team quality, and early evidence of execution. The compact rule does not reveal thresholds or relative weights. | Deep note.
• ★★★ | Is AI really going to force Indians out of jobs? — FCTC episode 6, part 1 | https://www.youtube.com/watch?v=Yzs7KJ5HQWw | Published/Event: 2024-12-10 | Format: long-form video interview / first of two parts; reliability: A for Prashanth Prakash's direct voice, B for exact wording in the third-party auto-generated caption mirror, and insufficient for medical forecasts or uncorroborated family and civic claims; access: partial; complete 1:22:52 third-party auto-caption mirror and publisher show notes inspected, but the official YouTube caption endpoint and yt-dlp returned a sign-in/bot challenge, creator-caption provenance was unavailable, and the audio was not independently checked | Unique value: The episode is principally a formative-biography and civic-operating interview, not an answer to its AI-and-jobs headline. Its sustained themes are Bengaluru identity, inherited access, family influences, lakes and urban systems, collaborative public action, and longevity; the only extended AI claim concerns health. | Deep note.
• ★★★ | How this VC straddles startups, politics and development — FCTC episode 6, part 2 | https://www.youtube.com/watch?v=eZUaJkNHCdo | Published/Event: 2024-12-17 | Format: long-form video interview / second of two parts; reliability: A for Prashanth Prakash's direct voice, B for exact wording in the third-party auto-generated caption mirror, and insufficient for rapid-fire deal economics or uncited market figures; access: partial; complete 56:35 third-party auto-caption mirror and publisher metadata inspected, but the official YouTube caption endpoint and yt-dlp returned a sign-in/bot challenge, creator-caption provenance was unavailable, and the audio was not independently checked | Unique value: Prakash opens the institutional history by rejecting a founder-hero account: he says Erasmic required three complementary co-founders and that Accel India would not have existed without that combination. The auto-captions garble one co-founder's name, so identity must come from independent records rather than silent correction. | Deep note.
• ★★★ | Future of Agritech: The evolution of Agritech in India | https://www.seedtoscale.com/podcast/future-of-agritech-the-evolution-of-agritech-in-india | Published/Event: 2020-12-11 | Format: 46:48 three-person panel podcast/video with publisher highlights; reliability: A for passages explicitly attributed to a named speaker; B for publisher synthesis; C for unsourced market sizes, penetration, and policy characterizations; access: partial — official audio/video and detailed page accessible, but no transcript or captions were obtainable | Unique value: The panel’s public agenda divides the market into landscape, funding tailwinds, farm bills, COVID effects, farm-to-consumer, precision agronomy, and founder advice. Mark Kahn, Prakash, and Anand Daniel all participate; only passages the page explicitly names can be assigned to one speaker. | Deep note.
• ★★ | Future of Beauty and Personal Care: Tracking the evolution of the category in India | https://www.seedtoscale.com/podcast/future-of-beauty-and-personal-care-tracking-the-evolution-of-the-category-in-india | Published/Event: 2021-08-06 | Format: 57:11 multi-speaker podcast/video with chaptered publisher highlights; reliability: A for direct speech if speaker-verified; B for publisher summaries and promotional framing; access: partial — official audio/video, RSS, and detailed chapter index accessible, but no transcript or captions were obtainable | Unique value: The publisher isolates Prakash’s category framing at 06:30–10:51, investor perspective at 20:50–24:00, and offline-survival point at 30:35–31:45. These ranges establish where to retrieve his views; the show notes do not provide their exact wording. | Deep note.
• ★★★ | Investments From US, Middle East Set To Make Up For Chinese Capital | https://www.newindianexpress.com/business/2020/Nov/22/interview--investments-from-us-middle-eastset-to-make-up-for-chinese-capitalprashanth-prakash-2226432.html | Published/Event: 2020-11-22 | Format: edited interview excerpts / Q&A; reliability: A for Prashanth Prakash's published answers, B for Bismah Malik's framing; access: full rendered excerpt; no audio, original transcript, or editing history | Unique value: Prakash described the first pandemic months as an observation period and estimated a six-month investment-market lag. By after August 2020, he believed investors had more evidence about which businesses fit the “new normal”; the interview supplies no market series or definition of recovery. | Deep note.
• ★★ | H-1B fee increase as a possible positive for Indian startups | https://www.linkedin.com/posts/cnbc-international_india-may-see-less-of-a-talent-drain-as-h-activity-7375846352586489856--JKF | Published/Event: 2025-09-22 11:00:07 UTC (creation time decoded from public activity ID) | Format: CNBC International video excerpt with public platform transcript; reliability: A for the visible speaker transcript; B for CNBC framing; access: full transcript for the short excerpt; linked full CNBC video route unresolved/robots-blocked | Unique value: Prakash viewed the announced H-1B fee increase as a potential positive for Indian venture and startups because some U.S.-experienced product and AI talent might return to India. This is a directional labor-supply hypothesis immediately after a policy announcement, not migration evidence. | Deep note.
• ★★ | We Can Do Much More For Our Country — India Philanthropy Summit 2026 fireside chat | https://www.youtube.com/watch?v=hLh-QuB7uBA | Published/Event: 2026-06-02 | Format: 20:15 HURUN INDIA fireside conversation between Prashanth Prakash and Anas Rahman Junaid; reliability: A for Prakash's direct words in HURUN INDIA's publisher clip transcript, B for the publisher's framing and role labels, and insufficient for the inaccessible remainder; access: partial; YouTube title/date/runtime metadata and one HURUN INDIA publisher clip transcript were inspected, but the full video has no available captions and direct audio retrieval was blocked by YouTube's sign-in challenge | Unique value: In the accessible excerpt, Prakash distinguishes emerging philanthropy from passive cheque-writing: he says the next generation wants direct involvement, institution-building, disruption, and scalable solutions. | Deep note.
• ★★★ | Borderless Ambitions — Globalising the India Stack | https://www.youtube.com/watch?v=kFzEr4c68sI | Published/Event: 2022-12-15 | Format: 48:21 India Global Forum upload containing a 20:13 Prashanth Prakash–Sanjay Jain panel followed by unrelated programming; reliability: A for official IGF metadata and session record; B for exact wording in public YouTube automatic captions; C for uncited company, market, and forecast figures; access: full English automatic captions through approximately 48:22; only the titled session at 00:03–20:13 is in scope, with moderator-resolved Prakash turns | Unique value: In a rapid opening exchange, Prakash answered that he had launched or funded about 300 startups, 25 of them unicorns. The question conflates launching, funding, personal, and firm activity, so the figures cannot be treated as a personal portfolio or performance denominator. | Deep note.
• ★★★ | The Start-up Pledge for Climate Action | https://www.youtube.com/watch?v=CE_DFKIrNYk | Published/Event: 2022-12-15 | Format: 23:12 multi-speaker India Global Forum panel; reliability: A for official IGF and ACT provenance; B for exact wording in public YouTube automatic captions; C for uncited comparative and forecast claims; access: full English automatic-caption track through 23:13 inspected; direct media retrieval was blocked, and only moderator-directed or otherwise turn-resolved Prashanth Prakash speech is used | Unique value: Prakash separated his Accel role from ACT, describing ACT as a coalition of venture funds, technology entrepreneurs, investment bankers, and social-sector leaders formed during COVID and then extended to climate, education, and healthcare. This is an institutional-origin account, not evidence of personal investment ownership or impact. | Deep note.
• ★★★ | Unboxing Bengaluru: The City of New Beginnings — India Global Forum | https://www.youtube.com/watch?v=Xykphi-RCBg | Published/Event: 2023-11-28 | Format: 19:23 one-on-one India Global Forum book conversation with Prashanth Prakash; reliability: A for direct speaker substance, B for exact wording in the third-party auto-generated caption mirror and IGF's publisher synthesis, and insufficient for unsourced city rankings or counts; access: partial; complete public auto-captions through 19:20, IGF's publisher summary, and YouTube title/date/runtime metadata were inspected, but creator-caption provenance was unavailable and audio was not independently checked | Unique value: Prakash frames Bengaluru as a bounded proxy for understanding India's last three to four decades of transformation. His subject is the city's social operating system—not technology alone—and why it repeatedly makes space for new institutions and entrepreneurs. | Deep note.
• ★★ | Accel's AI capital allocation and the IndiaAI Mission | https://www.linkedin.com/posts/accelinindia_the-indiaai-mission-aims-to-strengthen-the-activity-7338161302910996481-Kt17 | Published/Event: 2025-06-10 11:13:11 UTC (creation time decoded from the public activity ID); recorded at the 2025 Accel AI Summit | Format: short joint video interview with public LinkedIn transcript; reliability: A for identifiable speaker turns; B for auto-transcribed wording and Accel framing; access: full visible clip transcript; audio not independently checked | Unique value: Prakash says Accel remains primarily an early-stage investor—seed and Series A—with selective growth investments, and that this stage posture would continue for its AI-focused capital. The interviewer calls it a $650 million fund; the legal vehicle and AI-specific allocation are not established in the clip. | Deep note.
• ★★ | The Journey of Generosity at the 2026 Indiaspora Forum | https://www.youtube.com/watch?v=XRK7_PeHNJ8 | Published/Event: 2026-04-14 | Format: 35:06 four-person philanthropy panel; reliability: B for publisher metadata, participant roles, and event context; C for Prakash-specific content because no transcript or speech summary was recovered; access: partial — complete publisher description and event metadata recovered; transcript body and playable media unavailable | Unique value: Indiaspora explicitly lists Prakash, Arun Jain, and Latha Rajnikanth as panel speakers while separately identifying Atul Satija as session chair. That role separation establishes Prakash as a panel participant rather than host or introducer, although the inaccessible transcript prevents measuring his speaking share. | Deep note.
• ★★★ | How Infra.Market offers concrete solutions through its construction marketplace | https://www.seedtoscale.com/podcast/how-infra-market-offers-concrete-solutions-through-its-construction-marketplace | Published/Event: 2022-11-29 | Format: founder/investor podcast moderated by Pankaj Mishra; reliability: A for named first-person statements, B for exact wording in the indexed transcript mirror and publisher narrative, and insufficient for unaudited operating, valuation, and market-share figures; access: partial; official page, highlights, guest list, and public audio metadata were accessible, while a third-party indexed copy of the publisher-style transcript supplied the full speaker-labeled body; the official page did not expose a transcript, the mirror's original PDF route was Cloudflare-blocked, and no stable timestamps or page numbers were available | Unique value: Founder Souvik Sengupta frames the opportunity as a two-sided fragmentation problem: construction companies repeatedly rebuild local procurement networks, while small producers lack national distribution. His and Aaditya Sharda's complementary procurement and distribution experience is presented as the origin of the thesis. | Deep note.
• ★★★ | Grounding Innovation in Spirituality — INKtalks | https://www.youtube.com/watch?v=j8ViD-aqvYA | Published/Event: exact upload date unavailable | Format: 13:49 on-stage interview with Lakshmi Pratury / INKtalks; reliability: A for Prashanth Prakash's direct statements, B for exact wording in a third-party auto-generated caption mirror, and insufficient for geological, historical, event-scale, or city-wide claims; access: partial; the complete public auto-caption mirror was inspected, but the official YouTube page/captions returned an authorization or sign-in challenge, creator-caption provenance was unavailable, and audio was not independently checked | Unique value: Prakash frames the Nandi-area venue as a deliberate alternative to private consumption: a place for spiritual, creative, and immersive experiences rather than merely a large house. His ten-year vision is to make the area a spiritual and creative counterpart to Bengaluru's technology identity. | Deep note.
• ★★★ | Accel India founders on investing through instinct and intuition | https://yourstory.com/2020/01/investing-instinct-accel-india-investors-startups | Published/Event: 2020-01-16 | Format: edited joint interview article with embedded video; reliability: A for excerpts explicitly attributed to Prashanth Prakash or Subrata Mitra, B for framing and portfolio narrative; access: full rendered article; embedded joint video identified, but no full transcript or timestamp verification obtained | Unique value: Prakash's only substantive published excerpt says instinct and intuition must be used “in a guarded manner.” He warns that even deep familiarity with a space can create conditioning that becomes counterproductive, and connects the freedom to hold uncertainty with seed investing. | Deep note.
• ★★★ | We may be 5–7 years from creating ten $1-billion internet firms in India | https://www.business-standard.com/article/companies/we-may-be-5-7-years-from-creating-10-1-bn-internet-firms-in-india-prashanth-prakash-112032900025_1.html | Published/Event: 2012-03-29; page later shows 2013 update | Format: edited interview; reliability: A for attributed answers; B for framing; C for unsourced forecasts; access: indexed full text; direct page returned cache miss/empty body during this run | Unique value: Prakash’s 2012 internet thesis joins increasing broadband, consumer spending, and access: online catalogs could overcome the constrained assortment of physical retail in tier-II and tier-III cities. | Deep note.
• ★★★ | Invest in new ideas, not flavour of the season | https://www.livemint.com/companies/people/invest-in-new-ideas-not-flavour-of-the-season-says-accels-prashant-prakash-11703763460440.html | Published/Event: 2023-12-28 | Format: edited interview; reliability: A for attributed answers; B for reporter framing; C for unsupported forecasts; access: full text | Unique value: Prakash says Accel maintains “prepared minds,” reassessing spaces every four or five years and studying themes four or five years before expected impact. The investable target is a potentially large, globally relevant space with domestic tailwinds, not a familiar category attracting consensus capital. | Deep note.
• ★★★ | “We Expect 70–100 IPOs This Year,” Says Accel's Prashanth Prakash | https://www.financialexpress.com/business/sme-we-expect-70-100-ipos-this-year-says-accels-prashanth-prakash-3744114/ | Published/Event: 2025-02-10 | Format: edited interview excerpts / Q&A; reliability: A for Prashanth Prakash's published answers, B for Ayanti Bera's framing; access: full rendered excerpt; no recording, original transcript, or cited data appendix | Unique value: Prakash characterized India as an unusually complete emerging-market startup ecosystem because startups span sectors and reinforce capital and entrepreneurship. The comparison names Southeast Asia, South America, and China but gives no common variables, dates, or denominator. | Deep note.
• ★★★ | Karya AI and the Potential of AI to Uplift Bharat — Global AI Conclave | https://www.youtube.com/watch?v=-gVoLgI7lHI | Published/Event: 2023-12-21 video; panel held 2023-12-16 and reported 2023-12-18/19 | Format: 25:46 joint panel with Karya founder Manu Chopra and Prashanth Prakash; reliability: A for Prakash quotations in Moneycontrol's same-event report; B for reporter paraphrase and publisher chapters; access: substantial but not a transcript — official video metadata and chapters plus a same-event Moneycontrol report; captions/audio blocked | Unique value: Prakash argues that AI for lower-income and lower-literacy users needs simpler interfaces, especially voice interaction in Indian languages, because conventional search can be intimidating. His mechanism is interface and information synthesis, not merely cheaper compute. | Deep note.
• ★★★ | India's Startup Story at Khazanah Megatrends Forum 2022 | https://www.youtube.com/watch?v=XrQtEODMEeg | Published/Event: 2022-10-17 | Format: 24:28 single-speaker conference presentation within KMF2022 “Innovation Without Borders”; reliability: A for Khazanah Megatrends Forum metadata and session framing; B for exact wording in full third-party automatic captions; access: official session page plus captions inspected from 00:00 through approximately 24:22; direct YouTube replay and first-party captions were blocked | Unique value: Prakash divided India's technology-company formation into IT-services foundations, global product and engineering centers, and a startup/digitization phase. The chronology is a useful causal narrative but lacks cohort definitions, company counts by phase, and evidence separating correlation from causation. | Deep note.
• ★★★ | The kick-start fund | https://www.livemint.com/Companies/kk5qTJvCSaoWnHNYWUM68N/The-kickstart-fund.html | Published/Event: 2015-12-01 | Format: reported firm profile with edited Prashanth Prakash interview excerpts; reliability: A for attributed quotations, B for narrative, fund chronology, valuations, and deal figures; access: full rendered article; underlying interview and editing record unavailable | Unique value: Prakash described Accel's method as theme-based: select what it regards as the best team within a sector and “almost never” invest in conflicting companies. The absolute-sounding rule is immediately testable against Mint's later examples of portfolio consolidation and distress. | Deep note.
• ★★★ | India will see advanced-manufacturing unicorns within three to four years | https://www.moneycontrol.com/technology/india-will-see-manufacturing-unicorns-within-3-4-yrs-says-accel-s-prashanth-prakash-article-13397785.html | Published/Event: 2025-08-06 | Format: reported panel remarks at Accel's Advanced Manufacturing Summit; reliability: A for attributed direct quotations; B for reporter paraphrase and event framing; access: full rendered text | Unique value: Prakash said venture attitudes toward defence, robotics, semiconductors, aerospace, and other asset-heavy businesses had shifted in the preceding six to nine months. That is a participant's perception at an Accel-branded summit, not a measured industry-capital series. | Deep note.
• ★★★ | Markets Opportunities — Evaluate Wise & Execute Nice | https://www.seedtoscale.com/podcast/markets-opportunities-evaluate-wise-execute-nice | Published/Event: 2018-05-11 in the canonical RSS feed; page JSON-LD says 2018-03-18 | Format: 35:50 podcast with publisher highlights; reliability: A for passages explicitly attributed to Prashanth Prakash; B for SeedToScale summaries and examples; C for unsourced market-share or forecast claims; access: partial — official audio and extensive publisher highlights accessible, but no transcript or caption file was obtainable | Unique value: Prakash frames the first Indian consumer-internet opportunity as making access predictable: a venture must own the operational work required to deliver a consistently good customer experience, not merely add a digital interface. The publisher attributes this directly to him under “Take Ownership for Heavy Lifting,” but the wording was not checked against a transcript. | Deep note.
• ★★★ | Prashanth Prakash: Master mentor | https://www.livemint.com/mint-lounge/business-of-life/prashanth-prakash-accel-partners-vc-mint-business-lounge-profile-11715263874235.html | Published/Event: 2024-05-10 | Format: long-form reported profile/interview; reliability: A for attributed quotations; B for reporter narrative and retrospective claims; access: full text | Unique value: Prakash describes venture work as forming forward-looking views, locating entrepreneurs who can embody them, and serving as a guide whose involvement can range from operational detail to strategy when invited. This is a mentoring ideal, not evidence that every board relationship worked that way. | Deep note.
• ★★ | Medicine 4.0: Innovations for the Future of Indian Healthcare | https://www.youtube.com/watch?v=i8x_KMwf2io | Published/Event: 2025-09-25 | Format: 36:07 two-person fireside chat at RISE for Healthy Aging 2025; reliability: B for publisher description and organizer recap; C for speaker-level allocation within jointly summarized claims; access: substantial metadata and organizer recap but no recovered transcript or playable media | Unique value: Longevity India identifies Prakash and Nithin Kamath as the two fireside participants and says both unpack Medicine 3.0's infrastructure and personalization shortcomings. This establishes substantive co-discussion rather than a Prakash introduction or moderator-only role. Publisher description, paragraph 1 | Deep note.
• ★★★ | People are focusing on mobile-first or mobile-only approaches | https://economictimes.indiatimes.com/opinion/interviews/people-are-focusing-on-the-mobile-first-or-mobile-only-kind-of-approaches-subrata-mitra-prashanth-prakash-accel-partners/articleshow/48950203.cms | Published/Event: 2015-09-14 | Format: edited joint interview with Subrata Mitra and Prashanth Prakash; reliability: A for speaker-attributed answers; B for reporter framing; access: full rendered body | Unique value: Prakash says Accel did not foresee Indian e-commerce reaching its eventual scale when it first invested: consumer-adoption data were absent, connectivity was weak, and prior attempts had failed. This is unusually direct uncertainty evidence, though the joint article's introductory claim of who “spotted” Flipkart is not transaction-level attribution. | Deep note.
• ★★★ | More capital can compress advanced-manufacturing gestation periods | https://www.financialexpress.com/business/more-capitalcan-compress-gestation-periods-says-prashanth-prakash-3936416/ | Published/Event: 2025-08-05 | Format: edited Q&A with Ayanti Bera; reliability: A for Prakash's published answers; B/C for unsupported market estimates and causal forecasts; access: full rendered Q&A | Unique value: Prakash gives four reasons for Accel's advanced-manufacturing focus: China+1 supply-chain movement, a policy transition from production incentives toward design incentives, more engineering-first founder inbound, and a large market forecast. Each mechanism should be separately measured rather than accepted as one compound tailwind. | Deep note.
• ★★★ | From 500 to 2 Lakh Startups: How India Is Becoming a Global Tech Superpower | https://www.youtube.com/watch?v=Cbccg0VRBkc | Published/Event: 2026-03-04 | Format: MyGov India video interview with opening montage; reliability: A for official metadata; B for clear Prashanth Prakash answer turns and exact wording in full third-party automatic captions; B for interviewer framing only as interviewer claims; access: 1:16:04 runtime; captions inspected from 00:00 through approximately 1:16:01, but direct YouTube replay and first-party captions were blocked | Unique value: Prakash rejected a lone-investor version of Accel's successes, describing the named companies as firm and team investments. That is the correct attribution boundary for the rest of the interview unless a personal decision is expressly identified. | Deep note.
• ★★ | “No more e-commerce” — reported 2012 seed-funding pause | https://ventureintelligence.blogspot.com/2012/06/?m=0 | Published/Event: 2012-06-13 | Format: secondary trade-blog excerpt of a Business Standard interview; reliability: C for the excerpt and attribution; original publisher body not located; access: partial excerpt only | Unique value: A contemporary Venture Intelligence post reproduces Prakash saying Accel would stop seed-funding new e-commerce trading platforms because “me too” entry had crowded the space, while existing companies might scale, consolidate, or attract later-stage capital. | Deep note.
• ★★★ | There are not enough consumer brands; we have only seen the tip of the iceberg | https://www.moneycontrol.com/technology/there-are-not-enough-consumer-brands-weve-only-seen-the-tip-of-the-iceberg-accels-prashanth-prakash-and-fireside-ventures-kanwaljit-singh-article-12719055.html/amp | Published/Event: 2024-05-10 | Format: edited joint interview; reliability: A for speaker-labelled answers; B for reporter framing and co-authored report claims; C for forecasts; access: full text | Unique value: Prakash says consumer brands should scale through online and offline platforms “wisely and profitably”; the tipping point is brand formation, not permission to subsidize any omnichannel expansion. | Deep note.
• ★★★ | Pioneering Indian Venture Capital — Indian Silicon Valley | https://www.youtube.com/watch?v=OykUcA0GpYM | Published/Event: 2021-11-16 | Format: long-form video interview; reliability: A for Prashanth Prakash's direct voice, B for exact wording in the third-party auto-generated caption mirror, and insufficient for self-reported outcomes or personal economics; access: partial; complete 44:19 third-party auto-caption mirror inspected, but the official YouTube caption endpoint and yt-dlp both returned a sign-in/bot challenge, creator-provided caption provenance was unavailable, and the audio was not independently checked | Unique value: Prakash defines venture as a five-to-ten-year foresight exercise, but constrains the investor's role: a prepared mind should identify an emerging space and know what to look for, while the entrepreneur remains better placed to decide exactly what to build. | Deep note.
• ★★★ | Accel rejigs India playbook after previous misses | https://www.livemint.com/companies/accel-india-advanced-manufacturing-strategy-11754309450230.html | Published/Event: 2025-08-05 | Format: reported interview; reliability: A for attributed quotations; B for reporter framing and firm examples; C for unsourced market or employment claims; access: full text | Unique value: Prakash openly says Indian climate technology had not come close to its potential; Mint groups sustainability, agritech, and electric vehicles among Accel theses that underdelivered. The quotation supports climate tech directly, while the broader list is reporter framing. | Deep note.
• ★★★ | Indian startups need to deliver good profits in three to five years | https://www.business-standard.com/article/companies/indian-startups-need-to-deliver-good-profits-in-3-5-yrs-accel-s-prakash-122111601582_1.html | Published/Event: 2022-11-16 | Format: reported Bengaluru Tech Summit remarks; reliability: A for quoted remarks; B for reporter narrative; C for unsourced forecasts; access: indexed full text; direct publisher route may be paywalled | Unique value: Prakash distinguishes headline valuations from durable venture outcomes: good profits and public exits are necessary to return capital to investors and sustain later inflows. | Deep note.
• ★★★ | Shift towards profitability, not a funding winter | https://www.deccanherald.com/business/shift-towards-profitability-not-a-funding-winter-accel-india-founding-partner-prashanth-prakash-3858204 | Published/Event: 2026-01-10 | Format: edited interview; reliability: A for attributed answers; B for reporter framing; C for unsourced market totals and forecasts; access: full body embedded in page metadata; rendered extraction omitted body | Unique value: Prakash distinguishes a stage and sector reallocation from a generalized funding winter: he sees adequate $3–8 million pre-Series A and Series A capital in AI, deep tech, and some consumer segments, but a shortage in $20–50 million Series B/C rounds. | Deep note.
• ★★★ | India Can Become a Global Hub for IP-Led Innovation | https://www.youtube.com/watch?v=tyESNXlcBok and https://www.youtube.com/watch?v=N3q-eBrNuco | Published/Event: 2025-04-14 for tyESNXlcBok and syndicated PTI text; 2025-05-05 for N3q-eBrNuco | Format: one PTI interview appearance published as two 27:15 video uploads plus syndicated edited text; reliability: A for PTI's published answers and Prakash's contemporaneous endorsement; B for upload metadata and exact wording in third-party automatic captions; access: full syndicated text and near-complete captions through 27:16 inspected; YouTube replay/direct captions blocked, and the two uploads are consolidated as one appearance family | Unique value: Prakash said India could become a hub for IP-led innovation and high-value manufacturing in defence, electric vehicles, aerospace, and energy. He made this an opportunity thesis rather than documenting realized exports, patents, margins, or Accel returns. | Deep note.
• ★★★ | Demo Day 2024 keynote — Ramaiah Evolute | https://www.youtube.com/watch?v=eNX3GTiUq4o | Published/Event: exact upload date unavailable; title and event identify Demo Day 2024 | Format: 16:03 solo keynote; reliability: A for Prashanth Prakash's direct statements, B for exact wording in a third-party auto-generated caption mirror, and insufficient for market forecasts, portfolio counts, or future technical outcomes; access: partial; the complete public auto-caption mirror and YouTube's rendered title/channel metadata were inspected, but YouTube returned an authorization/sign-in challenge, creator-caption provenance was unavailable, and audio was not independently checked | Unique value: Prakash rejects unicorn size as the only valid outcome. He argues that a maturing public-company base creates buyers for smaller startups, provided those companies own deep technology, meaningful innovation, and globally relevant IP. | Deep note.
• ★★★ | Robotics, Precision Manufacturing Hot Bets For Investors | https://www.business-standard.com/technology/tech-news/robotics-precision-manufacturing-emerge-as-key-bets-accel-s-prakash-124112400535_1.html | Published/Event: 2024-11-24; page shows last updated 2024-11-25 | Format: reported interview / event article; reliability: A for Prashanth Prakash's exact quoted remarks, B for Peerzada Abrar's paraphrase, market figures, and context; access: full article body available; no recording, raw transcript, report dataset, or interview-question sequence | Unique value: Prakash identified robotics, precision manufacturing, and combinations of software with physical production as near-term founder and venture opportunities. He argued that the global integration gap and India's engineering and demand conditions created room for new entrants; the article does not define the gap or compare global competitors. | Deep note.
• ★★ | How to Start a Start-up and Digitalization of the Economy — Sahyadri Conclave 2018 | https://www.youtube.com/watch?v=0VtYhE9H_BU and https://www.youtube.com/watch?v=4jIQiI7-vZU | Published/Event: 2018-01-10 appearance date in Sahyadri's Day Five report; YouTube upload dates not exposed | Format: 69:29 two-part keynote/classroom talk (36:38 + 32:51), treated as one appearance; reliability: A for direct speech if speaker-verified; B for Sahyadri's contemporaneous institutional recap; access: partial — both official video records and a dated institutional recap accessible; no transcript, captions, chapters, or audio recovered | Unique value: Sahyadri's event report says Prakash's session explained how digitalization transformed industry, defined transformation and subsets of digitalization, and discussed data analytics, personalization, and network effects. This is the institution's near-event synthesis, not a verbatim transcript. | Deep note.
• ★★★ | A VC's Tips to Survive & Thrive in the Digital Age — Spotlight with Sandhya | https://www.youtube.com/watch?v=6pOe56BR7io | Published/Event: 2021-07-01 in the publisher RSS feed; YouTube premiere 2021-07-02 | Format: 22:45 one-to-one video interview; 22:39 publisher-hosted podcast audio; reliability: A for Prakash speech checked against the official audio; B for publisher title, description, and chapters; access: substantial — publisher RSS audio recovered and locally transcribed; public YouTube captions and video download were blocked | Unique value: Prakash treats the pandemic as an accelerator of existing digital marketplaces, payments, and fintech: changing customer behavior pulled some businesses roughly two years ahead of prior projections. This is a 2021 forecast made during the disruption, not a later explanation of winners. | Deep note.
• ★★ | Are Indian Startups Gaining from the New US Tariffs? — Startup Mahakumbh 2025 | https://www.youtube.com/watch?v=2YX_sVJFSRs | Published/Event: 2025-04-05 | Format: 8:53 one-to-one video interview; reliability: B for publisher metadata and description; C for summary-level speech attribution without a recovered transcript; access: partial — title, date, runtime, speaker identity, and publisher description recovered; public transcript and media routes were blocked | Unique value: TICE TV identifies Prakash as the interview subject and says he gives a “deep dive” into Startup Mahakumbh 2025's planning, purpose, first-day turnout, ecosystem implications, and next-year vision. That is sufficient to distinguish a substantive solo interview from a host introduction, although the inaccessible transcript limits claim resolution. | Deep note.
• ★★ | India's evolving ecosystem — SuperReturn Asia fireside | https://www.linkedin.com/posts/arnaudbonzom_it-was-a-pleasure-interviewing-prashanth-activity-7413716933432344576-D29m | Published/Event: 2026-01-04 23:04:17 UTC (creation time decoded from the public activity ID) | Format: interviewer and organizer summary of a SuperReturn Asia fireside; reliability: B for event identity and second-hand topic summary; C for unlinked numeric claims; access: detailed 21-point summary only; no recording or transcript published, and a public comment asks whether one will be released | Unique value: Interviewer Arnaud Bonzom and Black Mangroves summarize Prakash as arguing that India's ecosystem is approaching “completeness,” with domestic B2C IPOs adding to U.S.-listed SaaS precedents. Without a transcript, exact wording and valuation/exit criteria cannot be attributed directly. | Deep note.
• ★★ | Investments will follow good entrepreneurs in agritech — TechCircle interview | https://www.youtube.com/watch?v=1tBT3Dsvwv8 | Published/Event: 2020-11-26 | Format: 14:09 TechCircle interview with Prashanth Prakash at Bengaluru Tech Summit; reliability: A for the one direct quotation preserved by TechCircle, B for TechCircle's contemporaneous paraphrase and company/role metadata, and insufficient for the unusable auto-caption text; access: partial; YouTube title/date/runtime metadata and TechCircle's same-day publisher article were inspected, but the only public auto-caption mirror is severely garbled Hindi transliteration and direct audio retrieval was blocked by YouTube's sign-in challenge | Unique value: Prakash attributes agritech's improving opportunity set to smartphone adoption and asset-light technology companies tackling agricultural bottlenecks. This is TechCircle's contemporaneous paraphrase rather than a recoverable verbatim passage. | Deep note.
• ★★★ | The era of lead gen is game over; have full control of customer engagement | https://www.techcircle.in/2016/02/19/the-era-of-lead-gen-is-game-over-have-full-control-of-customer-engagement-prashanth-prakash/ | Published/Event: 2016-02-19 | Format: edited event interview article with a referenced video; reliability: A for attributed Prashanth Prakash quotations, B for framing and firm history; access: full rendered article; referenced full interview/video is not transcribed or timestamp-verified here | Unique value: Prakash expected entrepreneurs starting in 2016 to face more circumspect investors and a different fundraising reality, with a possible 12–24-month lull. He rejected the publication's bubble framing, treating bust concerns as a feature of hyper-valuation cycles. This was a contemporaneous forecast, not a timeless funding-cycle rule. | Deep note.
• ★★★ | TIA Summit 2023 — Unboxing Bengaluru and China-plus-N manufacturing | https://www.youtube.com/watch?v=PGF_n2ycEdE | Published/Event: exact upload date unavailable; recorded in the 2023 summit cycle, whose organizer concept note dates the event to 2023-11-21 and 2023-11-22 | Format: 28:58 on-stage interview; reliability: A for Prashanth Prakash's direct statements, B for exact wording in a third-party auto-generated caption mirror, and insufficient for self-reported fund scale, portfolio outcomes, city shares, or market forecasts; access: partial; the complete public auto-caption mirror and organizer event document were inspected, but official YouTube captions were unavailable behind an authorization/sign-in challenge, creator-caption provenance was unavailable, and audio was not independently checked | Unique value: Prakash treats Bengaluru as evidence that innovation has dispersed beyond Silicon Valley. He attributes the city's position to accumulated aerospace, space, IT/BT, startup, and digital-public-infrastructure capabilities, but his 60–80% share of consequential post-independence Indian activity is rhetorical and denominator-free. | Deep note.
• ★★★ | TIA Summit 2024 — domestic capital, frugality, and Global South platforms | https://www.youtube.com/watch?v=ap_IG2dTb_I | Published/Event: 2024-10-11; recorded 2024-09-25 according to the organizer agenda | Format: 17:39 fireside interview; reliability: A for Prashanth Prakash's direct statements, B for exact wording in a third-party auto-generated caption mirror, and insufficient for capital-flow figures, geopolitical assertions, market forecasts, or investment attribution; access: partial; the complete public auto-caption mirror, indexed YouTube metadata, and organizer agenda were inspected, but official YouTube captions were unavailable behind an authorization/sign-in challenge, creator-caption provenance was unavailable, and audio was not independently checked | Unique value: Prakash compares India's current startup phase with China's 2008–18 decade and argues that domestic capital—not foreign capital alone—is the step-change because it is persistent and locally committed. His claimed $4–5 billion monthly capital-market inflow is not sourced or directly tied to venture allocation. | Deep note.
• ★★★ | Value, Valuation & the Void in Between | https://www.youtube.com/watch?v=4qeABnsZtbE | Published/Event: recorded 2026-05-14; uploaded 2026-05-19 | Format: 39:25 TiE Bangalore D2C Summit fireside panel with audience Q&A; reliability: A for TiE metadata and participant/event record; B for exact wording in public YouTube automatic captions; C for uncited market, exit, and fund-size generalizations; access: full English caption track through approximately 38:53 inspected; direct media retrieval was blocked, and only resolved Prashanth Prakash turns are used | Unique value: Prakash treated direct-to-consumer as an entry point for building a brand around digital-first, trend-setting customers rather than a permanent channel definition. He placed the initial audience roughly at ages 25–35, but supplied no consumer data or category-specific exceptions. | Deep note.
• ★★ | The State of the VC Ecosystem amidst the COVID-19 Crisis | https://www.seedtoscale.com/podcast/the-state-of-the-vc-ecosystem-amidst-the-covid-19-crisis | Published/Event: 2020-04-17 in the canonical RSS feed | Format: 38:34 joint podcast/video with chaptered show notes; reliability: A for direct speech if speaker-verified; B for publisher summaries and chapter labels; access: partial — official audio/video and complete chapter index accessible, but no transcript or captions were obtainable | Unique value: The publisher structures the conversation from historical comparison (dot-com and 2008) through ecosystem response, portfolio risk, founder survival, pre-product-market-fit choices, a 24-month outlook, reinvention, sector change, and anxiety. This is reliable evidence of the agenda, not evidence that Prakash personally made each point. | Deep note.
• ★★★ | We Are Yet To See Slowdown In Business Plans We Receive | https://www.vccircle.com/we-are-yet-see-slowdown-business-plans-we-receive | Published/Event: 2008-11-14 | Format: edited email interview / Q&A; reliability: A for Prashanth Prakash's published answers, B for framing and vehicle description; access: full rendered Q&A; no audio or separate transcript applicable to an email interview | Unique value: Prakash said the successor fund would retain Erasmic's seed and early-stage emphasis while using Accel's information and relationship networks to help portfolio companies recruit, sell, and form strategic partnerships. The operating benefits were prospective claims, not measured outcomes. | Deep note.
• ★★ | Prashanth Prakash in Conversation at Venture Catalysts' Bangalore Office Launch | https://www.youtube.com/watch?v=YiuVUWpYi40 | Published/Event: 2023-08-06 | Format: 20:56 conversation with Dr. Apoorva Ranjan Sharma; reliability: A for official Venture Catalysts metadata; B for Sharma's contemporaneous promotional post and its Prakash-attributed excerpts; C for the uncited return figures in the video description; access: no first-party or automatic caption track and no retrievable media/transcript; note is limited to official show text and the interviewer's participant post | Unique value: Venture Catalysts' official title identifies a 20:56 conversation between Prakash and Dr. Apoorva Ranjan Sharma at the platform's Bangalore office launch. The title makes Prakash a materially substantive speaker, but the recording date is not stated separately from the August 6 upload. | Deep note.
• ★★ | Venture debt across the startup capital continuum | https://www.linkedin.com/posts/entrepreneur-media-india_entrepreneurindia-venturedebt-startupecosystem-activity-7315297655264268288-C7cu | Published/Event: 2025-04-08 09:01:12 UTC (creation time decoded from the public activity ID) | Format: short edited video interview at Stride Ventures' Global Venture Debt Report 2025 launch; reliability: A for identifiable Prakash answers; B for public auto-transcript and event framing; access: full visible clip transcript; audio not independently checked | Unique value: Prakash says Indian venture debt evolved from bridge/runway extension into working-capital, supply-chain, early bank-relationship, global-expansion, asset-based, and pre-IPO uses. This is a broad practitioner taxonomy presented at a lender's report launch, not comparative cost or default evidence. | Deep note.
• ★★ | Karnataka fintech, talent, and matched startup funding — World Fintech Summit 2025 | https://www.youtube.com/watch?v=sIv2_UQy9mA | Published/Event: 2025-05-20 | Format: 1:03:13 moderated session with Prashanth Prakash, Karnataka minister Priyank Kharge, and Razorpay co-founder Harshil Mathur; reliability: A for clearly attributable direct substance, B for exact wording and speaker boundaries in the third-party auto-generated caption mirror, and insufficient for unsourced ecosystem shares or forecasts; access: partial; public auto-captions through 1:03:01, Elets' publisher recap, and YouTube title/date/runtime metadata were inspected, but creator-caption provenance was unavailable and audio was not independently checked | Unique value: Prakash opens with a cluster-concentration thesis: he estimates Bengaluru/Karnataka companies account for more than 60% of payments and broking and roughly 50–60% of digital insurance, then argues the state should seek a larger share of national fintech. The shares are not sourced and may overlap definitions. | Deep note.
• ★★ | Ep #9: WTF Is Venture Capital? | https://www.youtube.com/watch?v=zCTm1wHcfkI | Published/Event: 2023-09-01 on YouTube; podcast editions dated 2023-09-02 | Format: edited five-person video/podcast roundtable; reliability: A for official Nikhil Kamath, Spotify, and Blume metadata; B for Karthik Reddy's editing recollection; no Prakash-content grade because speaker attribution is unresolved; access: full 2:29:40 runtime and unlabelled captions inspected; Glasp and a full third-party caption track merge speakers, and the recording could not be independently replayed in this run | Unique value: The official Nikhil Kamath title and Blume's participant page identify a venture-capital roundtable with Nikhil Kamath, Nithin Kamath, Rajan Anandan, Prashanth Prakash, and Karthik Reddy. Blume describes Prakash as a founding partner at Accel; that establishes participation and role, not ownership of any claim in the merged captions. | Deep note.
• ★★★ | Accel Partners: “We don't see us as investors, but co-creators” — YourStory | https://www.youtube.com/watch?v=tJLafC6zhbk | Published/Event: 2020-01-15 | Format: 36:07 joint video interview with Prashanth Prakash, Subrata Mitra, and Shradha Sharma; reliability: A for speaker-labelled excerpts in YourStory's companion articles, B for the publisher's synthesis of the full interview, and insufficient for statements not preserved in those articles; access: partial; YouTube's rendered metadata, the publisher's full-interview article and two companion excerpt articles were inspected, but YouTube exposed no captions, the public caption mirror confirmed no auto-captions, and yt-dlp audio retrieval failed with a sign-in/bot challenge | Unique value: YourStory frames Accel India's operating ethos as co-creation: work from the ground up with founders, add value across stages, and build trust rather than function only as a capital provider. This is publisher synthesis of the joint interview, not a speaker-labelled verbatim passage. | Deep note.

Biography, firm formation, and institutional history

• ★★★ | Padma Shri citation — Shri Prashanth Prakash | https://www.padmaawards.gov.in/Document/pdf/CitationsForTickets/2025/2025100.pdf | 2025 | Government of India award citation; reliability A for award/rationale, B for retrospective biography and portfolio claims; access full; locators: pp. 1–2, paras. 1–6 | High-value official identity/role narrative whose broad counts, chronology, and evaluative praise require independent transactional and contemporary controls.
• ★★★ | The alchemy of Bangalore | https://www.fortuneindia.com/technology/the-alchemy-of-bangalore/101014 | 2011 | long-form ecosystem history/interview; reliability A for direct quotes, B for narrative; access full; locator: Prakash/VisualReality/NetKraft passages | Period-proximate operator chronology, Atul Jalan/Kiran Nadkarni links, 1995/1998 dates, 2004 sale, and Bengaluru historical context.
• ★★★ | Erasmic, Accel merge to form India fund | https://www.business-standard.com/article/markets/erasmic-accel-merge-to-form-india-fund-108072901029_1.html | 2008-07-29 | contemporaneous financial press quoting announcement; reliability B/A for attributed firm statement; access full | $10 million Erasmic corpus, $60 million target, four partners, strategy, and named inherited portfolio at the merger date.
• ★★★ | Accel closes $60 million India fund in bearish market | https://www.livemint.com/Money/buVSNmUkp7oN7vr2fjTphJ/Accel-closes-60-mn-India-fund-in-bearish-market.html | 2008-11-13 | contemporaneous financial report/interview; reliability B; access full | Closing-date, mostly institutional LP base, Google non-repeat, sub-$1 million seed focus, and planned 25–30-company denominator; compare later $65 million wording.
• ★★★ | Sikshana Annual Report 2010–11 | https://sikshana.org/media/static/org/AnnualReports/AR2010_11.pdf | 2011 | nonprofit annual report; reliability A for contemporaneous Sikshana role/report, B for affiliated retrospective education/career biography; access full PDF; locators: pp. 3–4 and 33–34 | Contemporary chairman/board evidence plus a retrospective BIT/Delaware, S3/GSE/GemCase, and then-current-board account whose older claims still need independent records.
• ★★★ | US-based Adea Solutions to buy Netkraft | https://www.rediff.com/business/report/adea/20040610.htm | 2004-06-10 | PTI contemporaneous transaction report; reliability B; access full | All-cash acquisition with value undisclosed, $8.5 million raised, named investors, roughly 350 staff, and 2003–04 profitability; controls later exit-return storytelling.
• ★★★ | Still hot on start-ups | https://www.businesstoday.in/magazine/current/story/still-hot-on-startups-128574-2008-08-06 | 2008-08-24 print issue | contemporaneous Business Today firm profile; reliability B, with attributed Prakash comments; access full | Period-proximate account of Prakash and Subrata Mitra's five-deal investment club, later $10 million formal fund, named backing, and Accel combination; a stronger control than retrospective formation stories.
• ★★ | UnboxingBLR — Our Story | https://www.unboxingblr.com/our-story | current page accessed 2026-08-19 | official foundation history; reliability A for institutional self-description; access full; locator: Our Journey | Dates the 2020 project, 2022 book, 2023 foundation/institutional expansion, and later civic initiatives.
• ★★★ | Erasmic finds limited partner in Google | https://economictimes.indiatimes.com/industry/banking/finance/erasmic-finds-limited-partner-in-google/articleshow/1635402.cms | 2007-02-19 | contemporaneous business report with Prakash quotation; reliability B; access full | Rare near-contemporaneous evidence for Google's five-year LP commitment and Prakash's interpretation of it as institutional validation; amount and economics remain undisclosed.

Official Accel company relationship and denominator routes

• ★★★ | AgroStar | https://www.accel.com/companies/agrostar | current page opened 2026-08-19 | official Accel company page; reliability A for current partner/stage labels; access full | Named Prakash relationship and agritech case; connect to 2025 sector-level disappointment without assuming this company failed.

• ★★★ | BlueStone | https://www.accel.com/companies/bluestone | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Priority IPO case with prospectus and sole-authored thesis controls.

• ★★★ | BookMyShow | https://www.accel.com/companies/book-my-show | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Priority long-hold consumer case; page names Prakash and 2012 Series A.

• ★★★ | CleverTap | https://www.accel.com/companies/clevertap | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Priority SaaS/mobile-engagement case and possible WizRocket alias; deduplicate entity and rounds.

• ★★ | Deep Rooted | https://www.accel.com/companies/deep-rooted | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Perishables supply-chain case and denominator control.

• ★★ | FabHotels | https://www.accel.com/companies/fabhotels | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Travel/hospitality case requiring operating and outcome controls.

• ★★ | Fibmold | https://www.accel.com/companies/fibmold | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Advanced-manufacturing/sustainable-packaging case for strategy-evolution testing.

• ★★ | HomeLane | https://www.accel.com/companies/homelane | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Home-interiors case; compare with CapriCoast and avoid conflating follow-on/category overlap.

• ★★★ | Infra.Market | https://www.accel.com/companies/infra-market | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Priority B2B/building-materials case with SEBI/open-offer board and vehicle evidence.

• ★★★ | PropTiger | https://www.accel.com/companies/proptiger | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Named led-investment claim in older bios; reconstruct News Corp transaction, stage, board, and outcome.

• ★★★ | QwikCilver | https://www.accel.com/companies/qwikcilver | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Priority Pine Labs acquisition case; establish vehicle, cash/stock terms, and Prakash versus firm credit.

• ★★★ | RentoMojo | https://www.accel.com/companies/rentomojo | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Major rental-commerce relationship paired below with exact SEBI DRHP nominee-director evidence.

• ★★ | Snaptrude | https://www.accel.com/companies/snaptrude | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Architecture/design software lead and less-famous denominator case.

• ★★★ | Teabox | https://www.accel.com/companies/teabox | current page opened 2026-08-19 | official Accel company page; reliability A for current labels; access full | Early seed case with contemporaneous Prakash remarks and later operating-outcome questions.

• ★★ | Brik | https://www.accel.com/companies/brik | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Seed 2023; current website association only; all role and economic fields remain unknown.

• ★★ | CapriCoast | https://www.accel.com/companies/capricoast | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Seed 2015 and HomeLane acquisition label; transaction terms, role, and proceeds unknown.

• ★★ | Curefoods | https://www.accel.com/companies/curefoods | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Initial year 2021 with no displayed stage; current website association only.

• ★★ | EcoSoul | https://www.accel.com/companies/ecosoul | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Series A 2022; current website association and no realized economics.

• ★★ | Fixxly | https://www.accel.com/companies/fixxly | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Seed 2026; too recent for realized evaluation and current website association only.

• ★★ | Gently | https://www.accel.com/companies/gently | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Seed 2024; current website association only.

• ★★ | GlowRoad | https://www.accel.com/companies/glowroad | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Series A 2017 and Amazon acquisition label; consideration, ownership, and proceeds unknown.

• ★★ | MyWear | https://www.accel.com/companies/mywear | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Seed 2015; current website association only.

• ★★ | Pramaana Labs | https://www.accel.com/companies/pramaana-labs | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Seed 2026; too recent for realized evaluation and current website association only.

• ★★ | QikPod | https://www.accel.com/companies/qikpod | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Series A 2015; current website association with status and economics unresolved.

• ★★ | QTalk | https://www.accel.com/companies/qtalk | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Seed 2019; current website association with status and economics unresolved.

• ★★ | The Lafz | https://www.accel.com/companies/the-lafz | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Series A 2020; current website association only.

• ★★ | Travelstop | https://www.accel.com/companies/travelstop | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Seed 2019 and Wego acquisition label; terms, ownership, and proceeds unknown.

• ★★ | umrahme | https://www.accel.com/companies/umrahme | current page accessed 2026-08-20 | official Accel company page; reliability A for current partner/stage/outcome labels; access full | Series A 2017; deduplicated with HolidayMe through the official Traveazy lineage.

Transaction, vehicle, board, and economics controls

• ★★★ | Flipkart: Realizing the India Opportunity | https://www.accel.com/news/flipkart-realizing-the-india-opportunity | 2018-05-09 | official Accel firm account by Sameer Gandhi and Subrata Mitra; reliability A for firm statement, B for retrospective rationale; access full | Critical control against personal-credit inflation: $800,000 seed, $65 million fund wording, team process, and multi-vehicle financing without sole Prakash attribution.
• ★★★ | BlueStone Jewellery and Lifestyle Limited IPO prospectus | https://www.sebi.gov.in/sebi_data/attachdocs/aug-2025/1755172304134.pdf | 2025-08-14 | SEBI-hosted final prospectus; reliability A; access full, 532 pages; locators: cover, PDF pp. 24–25 and capital structure | Primary India III OFS share count, ₹517 price, ₹63.68 WACA, remaining shares, risks, and capital history.
• ★★★ | Swiggy Limited prospectus | https://www.swiggy.com/corporate/wp-content/uploads/2024/11/Swiggy-Limited-Prospectus_Final-filing-version.pdf | 2024-11-08 | company-filed prospectus; reliability A; access full PDF; locators: pp. 130–132 principal shareholders | Primary Accel India IV ownership evidence; does not assign Prakash sourcing, sponsorship, or personal economics, and Accel's company page instead identifies Anand Daniel.
• ★★★ | Shalimar Paints letter of offer | https://www.sebi.gov.in/sebi_data/commondocs/jan-2024/Shalimar_Paints_LOO_p.pdf | 2024-01 | SEBI-hosted offer document; reliability A; access full PDF; locators: sections 4.1.9–4.1.10 and director biography | Primary evidence that Prakash represented Accel India V on Infra.Market's board and for vehicle/entity separation.
• ★★★ | BookMyShow raises INR 550 crore | https://in.bmscdn.com/bmsin/static/pdf-press-release/6-feb-2017-bookmyshow-raises-largest-single-funding-of-inr-550crs.pdf | 2017-02-06 | company press release; reliability A for announced round and quotation, B for promotional claims; access full PDF | Contemporaneous quote and existing-investor role; control round chronology and co-investors.
• ★★★ | SEC Form D — Accel India IV L.P. | https://www.sec.gov/Archives/edgar/data/1635781/000163578115000001/xslFormDX01/primary_doc.xml | 2015-03-23 | US securities filing; reliability A; access full; locators: items 1, 3, 4, 13 and signature | Primary Cayman LP, $305 million offering, formation, and named-director evidence; listing Prakash is not proof of unilateral IC control, ownership, commitment, or carry.
• ★★★ | Freshworks Inc. IPO prospectus | https://www.sec.gov/Archives/edgar/data/1544522/000162828021019055/freshworks424b4.htm | 2021-09-21 | SEC-filed prospectus; reliability A; access full; locators: Principal Stockholders, pp. 148–149 | Security-specific counterevidence: the disclosed India III/IV Mauritius voting decisions were collective among four named people other than Prakash; does not resolve deal sourcing.
• ★★★ | SEC Form D — Accel India V L.P. | https://www.sec.gov/Archives/edgar/data/1687347/000168734716000001/xslFormDX01/primary_doc.xml | 2016-12-01 | US securities filing; reliability A; access full; locators items 1, 3, 13 and signature | Cayman vehicle and $430 million offering; preserve beside reported $450 million final close; no Prakash economics, performance, or complete IC authority.
• ★★★ | Deepening our longstanding commitment to India with our sixth fund | https://www.accel.com/news/deepening-our-longstanding-commitment-to-india-with-our-sixth-fund | 2019-12-02 | official Accel fund announcement; reliability A for firm statement, B for promotional history; access full | Announces a $550 million sixth India fund, seed/early focus, and more than 100 firm-level companies; not person-level capital or performance.
• ★★★ | Our commitment to companies in India and Southeast Asia | https://www.accel.com/news/our-commitment-to-companies-in-india-and-southeast-asia | 2022-03-01 | official Accel fund announcement; reliability A for firm statement, B for promotional statistics; access full | Announces a $650 million seventh fund and more than $2 billion regional commitment; firm-level only.
• ★★★ | Fueling the future: announcing Fund VIII for India’s boldest founders | https://www.accel.com/news/fueling-the-future-announcing-our-viii-fund-for-indias-boldest-founders | 2025-01-05 | official Accel fund announcement; reliability A for firm statement, B for promotional strategy; access full | Announces a $650 million eighth early-stage India fund and current sector map; too recent for realized-performance evaluation.
• ★★★ | BlueStone FY2024–25 annual report | https://bsmedia.business-standard.com/_media/bs/data/announcements/bse/05092025/20400aae-d730-4e9e-abb1-adb1cff066a9.pdf | 2025-09-05 | exchange-filed annual report; reliability A; access full PDF; locators: pp. 17–19 and 125–128 | Primary 2011 appointment, no personal beneficial holding, 17/17 attendance, no-debarment statement, current-directorship annexure, and India III/VII security-history control.
• ★★★ | RentoMojo draft red herring prospectus | https://www.sebi.gov.in/sebi_data/attachdocs/apr-2026/1775525404083_1204.pdf | 2026-04 | SEBI-filed DRHP; reliability A; access full PDF; locator: p. 235 and board/management section | Primary evidence that Prakash is a non-executive Accel nominee director; not proof of personal investment economics.
• ★★★ | Pine Labs — company history | https://www.pinelabs.com/about | current page accessed 2026-08-19 | acquirer official history; reliability A for company-stated transaction chronology; access full; locator: 2019 timeline | Independent company confirmation of the 2019 QwikCilver acquisition; price and investor proceeds are not disclosed.
• ★★ | Pine Labs acquires gift solution firm Qwikcilver for $110 million | https://www.livemint.com/companies/start-ups/pine-labs-acquires-gift-solution-firm-qwikcilver-for-110-mn-1555673014042.html | 2019-04-19 | PTI-reported acquisition announcement; reliability B; access full | Reports the completed $110 million acquisition and funding from internal cash plus existing investors; ownership, dilution, Accel proceeds, vehicle allocation, and Prakash economics remain unknown.
• ★★★ | Accel — Swiggy | https://www.accel.com/companies/swiggy | current page opened 2026-08-19 | official Accel company page; reliability A for current partner/stage labels; access full | Attribution counterexample: current page identifies Anand Daniel, not Prakash, for the 2015 seed relationship.
• ★★★ | Freshworks: a decade of aspirations and accomplishments | https://www.accel.com/news/freshworks-chennai-a-decade-of-aspirations-and-accomplishments | 2021-09-22 | official retrospective by Shekhar Kirani and Sameer Gandhi; reliability A for authorship/firm statement, B for retrospective narrative; access full | Attribution counterexample: foregrounds Kirani around the $1 million 2011 seed and early company work, not Prakash.

Government, policy, civic, legal, and contextual controls

• ★★★ | National Startup Advisory Council reconstituted; non-official members nominated | https://www.pib.gov.in/PressReleasePage.aspx?PRID=1987898&lang=2&reg=48 | 2023-12-18 | Press Information Bureau release; reliability A; access full | Primary appointment control naming Prakash and council mandate; does not document his individual recommendations.
• ★★★ | ACT Grants — About Us | https://actgrants.in/about-act/ | current page accessed 2026-08-19 | official institution page; reliability A/B for self-description; access full | Pandemic origin, venture-philanthropy model, Prakash board seat, and Health/Environment investment-committee roles; keep civic capital separate from Accel investing.
• ★★★ | ACT Capital Foundation Annual Return FY 2023–24 | https://actgrants.in/wp-content/uploads/2025/12/Annual-Return-FY-2023-24.pdf | FY ended 2024-03-31 | Companies Act annual return; reliability A; access full PDF; locators: pp. 8–11 | Primary evidence for promoter-member/director status, DIN 00041560, 25 shares, board attendance, nil remuneration, and nil penalties.
• ★★★ | IISc launches Longevity India Initiative | https://www.iisc.ac.in/events/iisc-launches-longevity-india-initiative-to-pioneer-ageing-research-in-india/ | 2024-04-18 | university release; reliability A for institutional act, B for promotional framing; access full | Confirms initial grant support and launch role; distinguishes research philanthropy from venture investment.
• ★★ | IVCA announces VC Council for 2024–2026 | https://static.levocdn.com/WGLBUT82/PressRelease-IVCAVCCouncil-2024-2026-hFB8vGgxIIGP.pdf | 2024-06-25 | IVCA press release; reliability A for council role, B for industry claims; access full PDF | Names Prakash and Rajan Anandan as co-chairs; maps industry-policy network and potential access advantages.
• ★★ | Karnataka Skill & Entrepreneurship Task Force report | https://www.karnatakadigital.in/wp-content/uploads/2023/04/Task-Force-Final-Report-10-03-2022-1.pdf | 2022-03-10 | Karnataka task-force report; reliability A; access full PDF; locator: member list and recommendations | Primary role evidence and policy/recommendation body for later public/private boundary analysis.
• ★★★ | Dinesh Kallahalli v. Prashanth Prakash — NGT judgment text | https://indiankanoon.org/doc/54167654/ | 2025-03-18 | secondary legal reproduction of National Green Tribunal judgment; reliability B pending canonical NGT comparison; access full copy | Gives allegations, respondent and regulator evidence, reasoning, and dismissal as devoid of merit; report allegations and disposition symmetrically and retain the official Supreme Court disposition.
• ★★★ | Dinesh Kallahalli v. Prashanth Prakash — Supreme Court order | https://api.sci.gov.in/supremecourt/2025/21995/21995_2025_7_32_61681_Order_09-May-2025.pdf | 2025-05-09 | official Supreme Court order; reliability A; access full, two pages; locator: signed order | Final no-interference disposition after Discovery Village environmental/planning litigation; allegations and dismissal must both be represented accurately.
• ★★ | Accel-backed Deep Rooted to shut operations | https://entrackr.com/exclusive/exclusive-accel-backed-deep-rooted-to-shut-operations-8865073 | 2025-03-18 | three-source reported shutdown; reliability C because sources are unnamed; access full | Strongest concrete Prakash-linked failure lead: reported shutdown after FY24 revenue fell 57% to ₹34.37 crore and loss reached ₹52.21 crore; use “reported,” not “verified.”
• ★★★ | SEBI Venture Capital Funds Regulations, 1996 | https://www.sebi.gov.in/acts/ventureregu.html | 1996 with later amendments | official regulation; reliability A; access full; locators: regulations 2–4, 11–21, 25–37 | Period-matched rules on registration, private capital pools, concentration, disclosure, investigation, and sanctions; do not assume every offshore Erasmic/Accel entity used this form.
• ★★ | India Venture Capital and Private Equity Report 2009 | https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID1579189_code1074013.pdf?abstractid=1579189&mirid=1 | 2009/2010 | IIT Madras academic report via SSRN; reliability B; access full alternate route | Quantitative 2004–08 context on boom/contraction, foreign-capital dominance, late-stage bias, and short holding periods; frame seed strategy without converting base rates into personal results.
• ★★★ | Inside Good Glamm’s collapse | https://economictimes.indiatimes.com/tech/startups/inside-good-glamms-collapse-a-timeline-of-how-the-crisis-unfolded/articleshow/122885143.cms | 2025-07-24 | reported timeline with named transaction events; reliability B; access full | Control against the 2021 co-authored MyGlamm thesis: flat round, defaults/arbitration, Accel board resignation, discounted asset sales, delayed salaries, and lender-driven breakup; exposure and responsibility remain team-level.

DEALS owner-phase additions and controls

• ★★★ | BlueStone 2011 original pitch deck | https://cdn.prod.website-files.com/6410c4927a4ae326761f2b26/68aef8bd41fb57631302dfed_Bluestone_Deck_PitchCraft.pdf | 2011; publisher copy accessed 2026-08-20 | company deck; reliability A for contemporaneous company claims; access full | Decision-freeze source for market, channel, inventory, margin, and expansion claims; not an Accel memo.

• ★★ | Deeksha — About Us | https://deekshalearning.com/about-us/ | current page accessed 2026-08-20 | official company history; reliability A for self-description, B for chronology; access full | Identity and operating-history control for Ace Creative Learning/Deeksha; no transaction economics.

• ★★★ | Rockefeller Foundation Form 990-PF 2019 | https://docs.candid.org/990/131/131659629/131659629_2019_202043119349100024_990PF.pdf | tax year 2019 | IRS-filed nonprofit return; reliability A; access full PDF; locator p. 31 | Rare LP-side fair values for India III–VI; values are not commitments, distributions, or returns.

• ★★ | Mahindra completes acquisition of BabyOye — reported outcome | https://economictimes.indiatimes.com/industry/cons-products/fmcg/mahindra-group-completes-the-acquisition-of-babyoye/articleshow/46115841.cms?from=mdr | 2015-02 | reported transaction; reliability B; access full | Characterizes a likely major VC write-off; exact Accel recovery and seller allocation remain unknown.

• ★★ | Venture capitalists’ role in starting a start-up — BabyOye | https://economictimes.indiatimes.com/jobs/venture-capitalists-role-in-starting-a-start-up/articleshow/8657278.cms | 2011-05-31 | contemporaneous founder report; reliability B; access full | Founder attributes warehousing, hiring, and operations help to Prakash after a $2.5 million round.

• ★★ | Snapdeal shuts down Exclusively.in | https://economictimes.indiatimes.com/snapdeal-shuts-down-exclusively/articleshow/53994968.cms?from=mdr | 2016-09 | reported shutdown; reliability B; access full | Post-acquisition closure control; does not establish investor recovery or zero value.

• ★★ | Accel, Prosus, Bessemer directors quit Good Glamm board | https://economictimes.indiatimes.com/tech/startups/accel-prosus-bessemer-directors-quit-good-glamm-board-amid-cash-crunch/articleshow/117679053.cms | 2025-01-29 | reported MCA/board account; reliability B; access full | Identifies Anand Daniel—not Prakash—as the available Accel board representative in the distress period.

• ★★ | Accel India to invest $2.5 million in HealthcareMagic | https://ehealth.eletsonline.com/2009/02/accel-india-to-invest-usd-2-5-million-in-healthcare-portal/ | 2009-02 | period financing report; reliability B; access full | Round-size and chronology source; seller allocation and later proceeds require acquirer filings.

• ★★ | Deep Rooted raises $12.5 million | https://entrackr.com/2022/09/d2c-startup-for-farm-produce-deep-rooted-raises-12-5-mn/ | 2022-09 | contemporaneous financing report; reliability B; access full | Follow-on decision freeze and comparable-category warning; no round allocation or board minutes.

• ★★ | FabHotels FY24 revenue and loss | https://entrackr.com/fintrackr/fabhotels-gross-revenue-crosses-rs-550-cr-in-fy24-losses-widen-23-8649057 | 2025-01 | reported ROC-derived financials; reliability B; access full | Scaled but loss-making ordinary-outcome control; not audited here and not a return statement.

• ★★ | Clover Ventures $5.5 million Series A | https://globalaginvesting.com/omnivore-accel-mayfield-back-5-5m-series-greenhouse-based-agtech-startup-clover-ventures/ | 2020-02 | contemporaneous financing report with Prakash quotation; reliability B; access full | Original perishables thesis and investor-attributed rationale before the pandemic channel break.

• ★★★ | IBBI insolvency process — Sanghvi Beauty & Technologies | https://ibbi.gov.in/claims/order-process/U72900PN2015PTC156163 | 2025 process | official insolvency registry; reliability A; access full | Current official CIRP control for Good Glamm legal disposition.

• ★★★ | NCLT order admitting Sanghvi Beauty & Technologies to CIRP | https://ibbi.gov.in/uploads/order/1347ac0abbba1338590531864725429d.pdf | 2025-09-12 | official tribunal order; reliability A; access full PDF; locators paras. 6–9, 15–28, 38–45 | Primary debt, default, operations-closed, inability-to-pay, and CIRP evidence; no Prakash conduct finding.

• ★★ | Teabox original seed announcement archive | https://in.teabox.com/blogs/press/india-s-teabox-raises-seed-funding-from-accel-to-build-an-online-starbucks-for-disrupting-global-tea-market | 2014 | company press archive; reliability B for announced thesis; access full | Period global-tea disintermediation thesis and logistics/technology execution boundary.

• ★★ | Teabox financials | https://inc42.com/company/teabox/financials/ | FY2024 | reported MCA-derived financial summary; reliability C/B; access full | Ordinary-outcome control suggesting small active profitability; no investor distributions.

• ★★ | Indiaspora member biography — Prashant Prakash | https://indiaspora.org/members/prashant-prakash/ | current page accessed 2026-08-20 | organization biography; reliability B/C for broad deal credit; access full | Claims foundational Ola/Freshworks roles; retained only as counter-attribution against transaction-specific sources.

• ★★★ | ITAT order — Prashanth Prakash land-and-villa capital-gains appeal | https://itat.gov.in/public/files/upload/-64755813590809135621351REFNO573-14-Prashanth_Prakash.pdf | assessment year 2009–10; pronounced 2015-03-20; accessed 2026-08-20 | official tribunal order; reliability A; access full PDF; locators paras. 2–7 and 14–18 | Concerns separate capital-gains treatment for land and a villa; allows the assessee's appeal and says the interest-deduction issue was not before the tribunal. It contains no Mu Sigma share sale or venture-return evidence.

• ★★★ | DPIIT FDI receipt table — FabHotels | https://master-dpiit.digifootprint.gov.in/static/uploads/2025/07/a291fdc82789fad0b2301e794a0a195c.pdf | government table accessed 2026-08-20 | official government PDF; reliability A; access full | Records ₹142.14 million from Accel India IV; not total cost, ownership, or return.

• ★★★ | QwikCilver: A First-to-Market Journey | https://medium.com/accel-india-insights/qwikcilver-a-first-to-market-journey-from-inception-to-acquisition-d3aa2e58a6d4 | 2019 | official Accel India retrospective; reliability A for firm statement, B for exit-era memory; access full | Joint Balachandran/Prakash founder-meeting account, category formation, and $20 million total-funding context.

• ★★ | Vedantu acquires Deeksha — reported $40 million | https://news.ventureintelligence.com/private-equity/edtech-unicorn-vedantu%EF%BB%BF-acquires-test-prep-platform-deeksha-for-%2440-m | 2022 | reported transaction; reliability B; access full | Company acquisition value only; seller allocation and Accel proceeds absent.

• ★★★ | Delaware pension minutes — Accel India VII | https://open.omb.delaware.gov/information/InvestCommittee/2022/02-14-22-DPERS-Minutes-FINAL-Public-signed.pdf | 2022-02-14 | public pension investment-committee minutes; reliability A; access full PDF | Approves up to $25 million to India VII subject to legal review; rare named-LP commitment evidence.

• ★★ | Exclusively.in Series A and board record | https://space.levo.so/WGLBUT82/7125574285168590105.pdf | period record accessed 2026-08-20 | archived deal PDF; reliability B; access full | Reports $2.8 million Series A and Prakash/Ashish Gupta board appointments.

• ★★ | Snapdeal acquires Exclusively.in | https://techcrunch.com/2015/02/18/india-snapdeal-exclusively/ | 2015-02-18 | contemporaneous transaction report; reliability B; access full | Reports acquisition after $18.8 million cumulative funding; terms and investor proceeds undisclosed.

• ★★ | Accel India V reported $450 million close | https://techcrunch.com/2016/11/30/accel-india-closes-its-newest-fund-with-450-million-and-in-record-time/ | 2016-11-30 | contemporaneous fund report; reliability B; access full | Final-close figure to preserve beside the $430 million Form D offering; no public bridge.

• ★★ | Teabox raises $7 million Series B | https://techcrunch.com/2017/12/13/teabox-raises-7m/ | 2017-12-13 | contemporaneous financing report; reliability B; access full | Follow-on and debt context for an ordinary outcome; no return data.

• ★★ | Deep Rooted financials | https://thekredible.com/company/deep-rooted/financials | FY2024 | reported ROC-derived financial summary; reliability C/B; access full | Revenue, expense, loss, and operating-cash-flow warning evidence; underlying filings not retrieved.

• ★★ | BabyOye buys Hoopos and raises fresh capital | https://timesofindia.indiatimes.com/tech-news/babyoye-buys-start-up-hoopos/articleshow/19483849.cms | 2013-04-09 | contemporaneous financing report; reliability B/C for ownership estimate; access full | Reports $12 million fresh capital and approximate combined investor ownership; not a cap table.

• ★★★ | Traveazy group timeline | https://traveazy.com/about/ | current page accessed 2026-08-20 | official company history; reliability A for self-described lineage; access full | Deduplicates HolidayMe and umrahme into one opportunity-group lineage.

• ★★★ | Accel — Acko | https://www.accel.com/companies/acko | current page accessed 2026-08-20 | official company page; reliability A for current partner label; access full | Anti-portfolio attribution control naming Abhinav Chaturvedi, not Prakash.

• ★★★ | Accel — Coverfox | https://www.accel.com/companies/coverfox | current page accessed 2026-08-20 | official company page; reliability A for current partner label; access full | Conflict/miss attribution control naming Subrata Mitra, not Prakash.

• ★★★ | Accel — Hotelogix | https://www.accel.com/companies/hotelogix | current page accessed 2026-08-20 | official company page; reliability A for current partner/stage labels; access full | Names Shekhar Kirani for the 2011 Series A; controls Prakash board association.

• ★★★ | Accel — Ola Cabs | https://www.accel.com/companies/ola-cabs | current page accessed 2026-08-20 | official company page; reliability A for current partner/stage labels; access full | Names Anand Daniel and Arun Mathew; counters broad Prakash foundational-role biography copy.

• ★★ | UrbanTouch tribunal record | https://www.casemine.com/judgement/in/5f8555a7714d58330c70a4ca | judgment reproduction accessed 2026-08-20 | legal-document reproduction; reliability B pending canonical tribunal PDF; access full | Identifies Accel India Venture II (Mauritius), ₹58,752,280 invested across equity/Series A/Series B, ₹35,017,255 sale consideration, losses and shutdown; supports only a single-source observed ~0.596x sale-consideration / cited-investment ratio before fees, taxes, timing, and unobserved distributions; it is not a fund return or Prakash economics.

• ★★★ | CIMA private-fund register — June 2022 | https://www.cima.ky/upimages/commonfiles/QuarterlyListofallPrivateFundsregisteredandlicensedwiththeCaymanIslandsMonetaryAuthority-June2022_1657640012.pdf | 2022-06 | official regulator register; reliability A; access full PDF; locator p. 3 | Lists India VII main and Entrepreneurs LP separately; registration is not a performance statement.

• ★★ | IDG/EVF investment report and partner biographies | https://www.ciol.com/idg-ventures-india-erasmic-venture-fund-invest-usd35-perfint/ | period report accessed 2026-08-20 | contemporaneous industry report; reliability B; access full | Names Subrata Mitra, not Prakash, for Small Device/Mu Sigma governance controls.

• ★★ | Clover uprooted to Deep Rooted | https://www.forbesindia.com/article/how-we-survived/clover-uprooted-to-deep-rooted/67291/1 | 2021 | reported company survival account; reliability B; access full | Documents the pandemic demand collapse and channel pivot; no investor decision record.

• ★★ | Lastaki advises ZingHR | https://www.linkedin.com/pulse/lastaki-advises-zinghr-concluding-investment-from-neha | transaction period; accessed 2026-08-20 | adviser account; reliability B/C; access full | Associates ZingHR/Cnergyis with Accel and Mumbai Angels; no exit or economics.

• ★★ | FabHotels raises $25 million | https://www.livemint.com/Companies/m9ZwLqVyQmf44QKjgLqTuK/FabHotels-raises-25-million-from-Goldman-Sachs-Accel-Partn.html?facet=amp | 2017 | contemporaneous financing report; reliability B; access full | Round chronology and follow-on participation; ownership and return absent.

• ★★★ | Mahindra acquisition disclosure — BabyOye | https://www.mahindra.com/resources/investor-reports/FY17/Announcements/SEintimationMRPL14thOct.pdf | 2015-02; filed copy accessed 2026-08-20 | official company disclosure; reliability A; access full PDF | Confirms acquisition; does not disclose price or investor recovery.

• ★★ | BabyOye funding and thesis | https://www.medianama.com/2011/04/223-babyoye-funding-accel-tiger-global/ | 2011-04 | contemporaneous financing report; reliability B; access full | Decision-date category thesis and $2.5 million round evidence.

• ★★★ | Accel India III $155 million official close | https://www.prnewswire.com/news-releases/accel-partners-announces-closing-of-155-million-seed-and-early-stage-fund-for-india-133962338.html | 2011-11-15 | official Accel release; reliability A for firm statement, B for promotional portfolio claims; access full | Fund size, strategy, 34-company team denominator, and portfolio examples; not Prakash-specific allocation.

• ★★ | Digital Chocolate expands to India and Spain | https://www.rcrwireless.com/20070813/archived-articles/digital-chocolate-expands-to-india-spain | 2007-08-13 | contemporaneous acquisition report; reliability B; access full | Small Device Technologies acquisition control; transaction timing differs from archive shorthand.

• ★★★ | SEC Form D — Accel India VI L.P. | https://www.sec.gov/Archives/edgar/data/1784044/000178404419000001/xslFormDX01/primary_doc.xml | 2019-11-27 | US securities filing; reliability A; access full; locators items 1, 3, 13–14 | Records $525 million offered/sold to 96 investors; preserve beside Accel $550 million announcement.

• ★★★ | SEC Form D — Accel India VII Entrepreneurs L.P. | https://www.sec.gov/Archives/edgar/data/1913365/000191336522000001/xslFormDX08/primary_doc.xml | 2022-03-18 | US securities filing; reliability A; access full; locators items 1, 3, 13–14 | Records separate $49.42 million entrepreneur sleeve sold to 149 investors; do not add mechanically to headline.

• ★★★ | SEC Form D — Accel India VIII L.P. | https://www.sec.gov/Archives/edgar/data/2044369/000204436924000001/xslFormDX01/primary_doc.xml | 2024-12-20 | US securities filing; reliability A; access full; locators items 1, 3, 13–14 | Delaware LP, $650 million sold to 131 investors, and related-person control.

• ★★★ | NCR 2005 Form 10-K | https://www.sec.gov/Archives/edgar/data/70866/000119312506051223/d10k.htm | 2006 filing for FY2005 | SEC-filed annual report; reliability A; access full | Independent Galvanon asset-acquisition control; Erasmic exposure and proceeds remain unproved.

• ★★★ | Ebix 2014 Form 10-K | https://www.sec.gov/Archives/edgar/data/814549/000081454915000012/ebix-2014x10k.htm | 2015-03-16 | SEC-filed annual report; reliability A; access full; acquisition note | HealthcareMagic $6 million cash plus contingent earnout up to $12.36 million; no seller waterfall.

• ★★★ | Ebix 2016 Form 10-K | https://www.sec.gov/Archives/edgar/data/814549/000081454917000004/ebix-2016x10k.htm | 2017-03-16 | SEC-filed annual report; reliability A; access full; acquisition note | Later HealthcareMagic earnout-liability control showing zero carrying amount.

• ★★ | Catamaran and Accel invest ₹5 crore in Ace Creative Learning | https://www.vccircle.com/catamaran-accel-investing-rs-5cr-test-prep-firm-ace-learning | transaction-period report accessed 2026-08-20 | deal report quoting Prakash; reliability B; access full | Prakash confirms Accel’s ₹2.5 crore commitment; stake and security undisclosed.

• ★★ | Hotelogix gets $1.2 million from Accel and others | https://www.vccircle.com/hospitality-saas-solutions-firm-gets-12m-accel-partners-others | 2011 | contemporaneous financing report; reliability B; access full | Quotes Shekhar Kirani and supplies the stronger transaction attribution.

• ★★ | Zauba DIN profile — Prashanth Prakash | https://www.zaubacorp.com/PRASHANTH-PRAKASH-00041560 | current mirror accessed 2026-08-20 | commercial corporate-registry mirror; reliability C pending MCA filings; access full | 71-entity lead list for board chronology; explicitly not an investment denominator.

• ★★ | Deep Rooted pandemic pivot | https://yourstory.com/2021/04/pandemic-survival-bengaluru-farm-to-fork-d2c-startup-deep-rooted-co | 2021-04 | reported founder/company account; reliability B; access full | Contemporaneous capital-intensity, trust, logistics, affordability, and two-city warning evidence.

• ★★ | VilCart company profile and funding history | https://yourstory.com/companies/vilcart | current profile accessed 2026-08-20 | publisher company profile; reliability B/C; access full | Names Prakash among 2021 seed investors; vehicle, check, ownership, and economics unknown.

• ★★ | Zauba company profile — Clover Ventures | https://www.zaubacorp.com/CLOVER-VENTURES-PRIVATE-LIMITED-U74999KA2018PTC112837 | current mirror accessed 2026-08-20 | commercial corporate-registry mirror; reliability C pending MCA filings; access full | Active-entity and nominee-director lead; does not prove operating activity, current tenure, votes, or investment economics.

Coverage assessment

Mechanical coverage

This map contains 220 rows and 220 unique canonical source-family URL fields: 136 ★★★ / 84 ★★ / 0 ★. Exclusive section totals are 26 direct writings/first-person institutional records; 74 interviews, talks, video, or edited-interview families; 9 biography/firm-history records; 28 official Accel company routes; 17 transaction, vehicle, board, or attribution controls; 12 government, policy, civic, legal, or contextual controls; and 54 DEALS owner-phase additions and controls. The map is 215 checked / 5 unchecked: all writing, spoken, and DEALS sources used in their owner phases are checked, while five unused biography/civic context rows remain preserved for later synthesis or review. Multi-upload PTI and Sahyadri appearances remain intentional family rows rather than duplicates. Mirrors, transcript tools, filing indexes, author indexes, search-result pages, short clips, and trailers are not counted independently.

Identity, guiding questions, and name/date sweep

The SCOUT was framed around eight questions: (1) exact identity and chronology; (2) the operator-to-investor transition; (3) Erasmic/Accel entities and decision rights; (4) the complete written and spoken corpus; (5) person-versus-team-versus-vehicle deal attribution; (6) failures, criticism, litigation, and regulatory evidence; (7) the boundary among policy, civic, family/personal, and Accel roles; and (8) the missing portfolio and return denominator.

Searches used Prashanth Prakash, Prashant Prakash, Prakash Prashanth, Visual Reality, VisualReality, NetKraft, Netkraft, Erasmic, Accel India, partner/company combinations, year sweeps from the 1990s through 2026, and criticism/lawsuit/NGT/Supreme Court/SEBI/MCA variants. Government, Accel, publisher, corporate-filing, and biographical sources tie the target to Bengaluru, the University of Delaware, Visual Reality, NetKraft, Erasmic, Accel, DIN 00041560-linked board records, Karnataka policy roles, and named civic institutions. Namesake legal and corporate results were excluded unless another target-specific identifier connected them.

Direct voice, writings, and spoken-source denominator

The priority written denominator now contains twenty-six records: nine 2008 Erasmic institutional pages recovered from Common Crawl; the Accel first-person profile; the BlueStone, MyGlamm, and Captain Fresh investment essays; the 2020 Accel India history; five clear sole-bylined 2023–24 policy/manufacturing op-eds plus one 2022 Prakash/TNN-credited civic essay whose footer identifies Prakash as writer; the 2024 Decoding Omnichannel and 2025 Advanced Manufacturing Leap reports; the signed Bengaluru Innovation Report 2024 foreword; Unboxing Bengaluru; and two direct 2025–26 LinkedIn writings. Twenty-four bodies are publicly readable in full, while the omnichannel report and book are partial but substantial. The Erasmic pages are contemporaneous firm/team voice rather than Prakash-personal writings; exact capture and ARC metadata are preserved in their notes. Old Accel India, wider LinkedIn/X, conference-publication, policy-submission, and remaining book routes were also searched before fixing this denominator.

The TALKS denominator is now 74 processed appearance families, deduplicated at the appearance level and linked one-to-one to 74 distinct notes/t-*.md deep notes. All 74 registry rows are checked. Edited interviews and reported remarks count when they preserve attributable Prakash answers; multi-part or duplicate uploads count once when they represent the same appearance. FCTC parts 1 and 2 remain separate because they contain distinct interview bodies, while the two PTI uploads and two Sahyadri parts remain consolidated family rows.

Access is heterogeneous and is stated row by row. On a conservative binary classification, 38 families are full and 36 are partial/limited: “full” requires an Access field that explicitly begins full/indexed full or documents captions spanning essentially the entire runtime; fields marked partial, substantial, metadata/summary-only, or no transcript are counted partial. The underlying routes include full edited text; first-party audio checked locally; full public automatic captions; complete third-party auto-caption mirrors where first-party captions were blocked; publisher transcripts, chapters, or same-event reporting; and, for the weakest retained families, metadata or publisher summaries only. Third-party captions are never described as creator captions, audio was not claimed checked unless it was, joint-speaker claims remain speaker-bounded, and metadata-only families carry no invented transcript substance. Repetition across appearances is consistency evidence, not proof of accuracy, attribution, or investment efficacy.

Explicit exclusions do not inflate the 74-family denominator. Self Care the Foundation for Success (uploaded 2025-09-22; 61:40) names Prakash in a three-person program, but the inspected public evidence did not establish a sufficiently material, attributable Prakash contribution. Biopeak Live episode 11 with Matt Kaeberlein (uploaded 2026-05-25; 57:46) contains Prakash host questions and transitions; Kaeberlein supplies the scientific, clinical, and entrepreneurial claims, and no Prakash investment or career decision evidence appears. Short promotional clips, trailers, excerpt reposts, and duplicate uploads were either consolidated into their fuller appearance family or excluded; none received a standalone row merely because Prakash's name appeared in the title or metadata.

Vehicle, people, and attribution boundary

The mapped public sequence now distinguishes: the reported $10 million Erasmic corpus; the $60 million 2008 close versus Accel's later $65 million wording; India III's official $155 million close; India IV's $305 million SEC offering; India V's $430 million SEC offering versus a reported $450 million final close; India VI's $525 million Form D sale versus Accel's $550 million announcement; and $650 million India VII and VIII announcements/filings. These records establish different facts—close, offering, vehicle-family scope, or firm announcement—not a continuous audited performance series. Management/advisory agreements, LP agreements, allocations, carry, IC records, cash flows, and Prakash-personal economics remain unavailable.

Flipkart is the central attribution warning. Accel's own 2018 account is authored by Sameer Gandhi and Subrata Mitra and describes team and multi-vehicle support. The Padma citation and promotional biographies group Flipkart among Prakash's bets without transaction-level role evidence. Until a contemporaneous memo, allocation, board record, or participant account resolves it, the roster's “seeded Flipkart” shorthand is not treated as personal lead credit.

Candidate portfolio denominator

Accel's current profile exposes 28 company affiliations in structured data and relationship cards; all linked pages were opened and reciprocally name Prakash. DEALS expanded that mutable graph to 52 deduplicated opportunity groups by adding named historical EVF/Erasmic cases, Padma noncurrent cases, personal/non-Accel-looking investments, shared essay ties, board/transaction cases, and team/profile attribution controls. Fourteen current company pages remain standalone priority rows and the other 14 are traceable through the master profile graph; the 54 owner-phase additions map every new filing and outcome source used in investments.md.

The searched state is 52/52 named candidates researched, 6/52 developed as decision labs, 0/52 sufficiently reconstructed end to end, and 52/52 unresolved in at least one material attribution or economic field. The Padma citation first names eight bets and then says “over 50 other companies,” without a roster or definition; Erasmic's reported 15-company total is only partially named. A career coverage fraction and literal Pareto-return split therefore remain not computable. [researcher inference · strong] DEALS denominator

Criticism, litigation, misses, and skeptical searches

Mandatory searches combined Prakash/Accel/Erasmic/portfolio names with criticism, bad behavior, conflict, underperformance, failed thesis, loss, write-off, lawsuit, fraud, SEBI, NGT, Supreme Court, Discovery Village, board, and government-adviser variants. DEALS added the UrbanTouch tribunal record, BabyOye's reported likely write-off, Exclusively.in's post-acquisition shutdown, Deep Rooted's reported operating stop, Good Glamm's official CIRP, the Policybazaar conflict miss, and Prakash's climate-tech underdelivery admission. The NGT dismissed the Discovery Village application and the Supreme Court declined to interfere. No inspected source established a securities sanction, fraud finding, or investment-related merits judgment against Prakash. That is a bounded search result, not proof no such record exists.

Access barriers and archive routes

• Unboxing Bengaluru was partially read through authorized Penguin, Moneycontrol, and Times of India excerpts; obtain a licensed copy or legitimate library loan for complete chapter, notes, bibliography, and co-author analysis.
• The NLSIU catalogue identifies Malini Goyal and Prashanth Prakash, Vintage/Penguin Random House, ISBN 9780670098521, xvii + 331 pages, call number 307.760954872 GOY, and one General Stacks copy marked available; the body was not borrowed or read in this run. NLSIU catalogue route
• All 74 TALKS appearance families were processed through the transcript hierarchy and carry an exact row-level access statement. First-party audio or creator captions remain unavailable for a material subset; those notes preserve the actual fallback used—public automatic captions, third-party caption mirrors, publisher transcripts/chapters, same-event reporting, or metadata-only evidence—and do not claim audio verification or full-speech coverage.
• MCA master data, LLP agreements, annual returns, and charge/ownership documents require direct MCA retrieval; commercial registry mirrors are leads only and were not admitted as independent source rows.
• Internet Archive CDX attempts for erasmic.com/*, net-kraft.com/*, and visualreality.com/* each timed out after 30 seconds. A subsequent Common Crawl 2008–09 index/ARC pass recovered nine full Erasmic pages with exact capture metadata; NetKraft remains tied only to contemporaneous reporting, and the Visual Reality historic-domain identity remains unconfirmed. Wayback failure is therefore an access-route result, not an Erasmic corpus gap.
• Venue sweeps found no target-attributable item at Stanford eCorner, HBS oral histories, 20VC, Acquired, Colossus, 25iq, The Knowledge Project, Tim Ferriss, TED, C-SPAN, or Charlie Rose. This is a bounded negative search, not proof that no appearance exists.
• LinkedIn/X, old Erasmic/Accel India, Bengaluru Tech Summit, IVCA, Startup Mahakumbh, policy-submission, and conference combinations were searched. Two full-body LinkedIn posts were recovered and processed, with dates decoded from their public activity IDs; no stable complete social corpus or attributable policy submission surfaced. Conference results duplicated listed media or supplied metadata-only leads.
• Forbes India returned HTTP 403; the IIT Madras copy of the 2009 VCPE report triggered CAPTCHA, so the full SSRN route was used; direct KDEM timed out; and one PIB iframe/cache route failed before a canonical PressReleasePage.aspx result supplied the indexed content. Claims dependent only on failed routes were excluded.
• The full NGT judgment text was inspected through a legal-document reproduction and the Supreme Court disposition is official; a stable canonical NGT judgment PDF and the complete pleading set still need retrieval.
• Private fund LP reports, capital calls/distributions, allocation schedules, IC memos, board minutes, and complete pass/write-off records did not surface publicly.

Search saturation

After the source-category, venue, social, catalog, archive, vehicle, filing, and adverse sweeps above, the closing saturation pass used three distinct unresolved queries: the canonical NGT final-order PDF for OA 49/2023; Prakash-specific Accel India IV carry/IC/allocation records; and Prakash-linked Deep Rooted/Good Glamm loss or write-off records. All three returned no new result. This supports closing public-route discovery for SCOUT; it does not claim private records, archive bodies, books, or media transcripts are complete, and their exact owner-phase handoffs remain listed below.

DEALS then reconciled the 52-group union and closed with three distinct transaction searches: Prakash-specific UrbanTouch/Accel consideration; HealthcareMagic/Accel earnout allocation; and Deeksha/Accel/Vedantu proceeds. They returned no additional ownership, IC, or distribution evidence beyond the tribunal, acquirer filings, and reports mapped above. This closes the public-route DEALS sweep, not private fund, MCA, board, or closing records.

WRITINGS then ran three owner-phase closing sweeps: (1) the Moneycontrol and Times of India author indexes plus exact-name/date variants, which surfaced the three sole-bylined op-eds and one Prakash/TNN-credited civic essay now processed and no further priority body; (2) Erasmic/Accel India wildcard searches across Wayback and Common Crawl, which recovered the nine processed Erasmic pages while Wayback remained unavailable; and (3) identity-verified LinkedIn/X searches around manufacturing, startup policy, IP, and Bengaluru, which produced the two processed canonical LinkedIn posts but no complete stable social feed or attributable policy submission. Two later profile snippets lacked stable canonical post URLs and were not treated as bodies. No identified accessible ★★★/★★ writing remains unprocessed.

Historical context and completed-peer boundary

SEBI's period regulation, the 2009 IIT Madras venture-capital study, Karnataka's task-force report, IVCA, contemporaneous 2008 fund reports, and Prakash's dated views provide starting points for India's seed-capital, regulation, public-market, talent, and policy context. Later phases must add period-matched quantitative base rates without treating current datasets as ex-ante facts.

No completed canon investor matches Prakash's modern India geography and era. Peter Brooke is a limited structural comparator for building an investment institution across markets, while Georges Doriot offers a limited patient-capital/public-policy contrast. Neither is a direct strategy peer; unfinished India/modern investors must not be used as completed comparators.

Highest-value retrievals for later phases

1. Obtain a licensed full copy of Unboxing Bengaluru, a stable full Decoding Omnichannel download and underlying data, additional old Accel India archive bodies, a more complete identity-verified social feed, and attributable policy submissions; the nine located Erasmic pages and all other accessible priority writings are already processed.
2. Recover first-party audio or creator captions for the partial-access TALKS families, especially metadata- or publisher-summary-only appearances; preserve the completed family denominator and joint-speaker attribution rather than adding duplicate clips.
3. Obtain Erasmic and Accel formation, LP, management/advisory, IC, allocation, and cash-flow records; reconcile the $60/$65 million vehicle descriptions and Google commitment.
4. Build the career deal ledger beyond the 28 live-profile links, including historical removals, firm-only claims, personal investments, passes, write-offs, sector disappointments, and ordinary outcomes.
5. For Flipkart, Swiggy, BlueStone, BookMyShow, Infra.Market, QwikCilver, and other material cases, retrieve contemporaneous memos, filings, founder/partner accounts, board records, ownership, dilution, and cash flows before assigning credit or economics.
6. Retrieve MCA primary records, a canonical NGT judgment file, and the complete pleadings; map policy, nonprofit, personal/family, and Accel entities with conflict and recusal boundaries.
7. Add India venture-capital, public-market, mobile/internet, regulation, and sector base rates matched to each decision date; label every later reconstruction.

Source notes