Fundraising for AI products

Study exceptional raises, learn the process, and build an investment case grounded in customer value.

Founder testimony, investor reasoning, practical fundraising craft and selected primary financing disclosures. The round table is a set of case studies, not a complete funding history or a ranking of current companies. Large raises do not by themselves reveal the process that caused them.

14People
50Catalogued sources
50Total sources
6Practical playbooks

01 / Start here

The sources to open first

Aravind Srinivas

Perplexity AI CEO Aravind Srinivas, PhD 21, on why he ditched pitch decks

2025-10-21 · Founder talk with primary transcript

Srinivas explains memo-led investor diligence and follow-up questions.

Open source ↗ · Evidence notes

May Habib

WRITER CEO May Habib on Full-Stack Enterprise AI

2025-11-06 · Founder masterclass recap

Habib uses fundraising to align the team on a multi-year story and stress-test convictions through repeated investor questioning.

Open source ↗ · Evidence notes

Michael Seibel

How to Pitch Your Company

2016-07-19 · Article

A seven-question pitch: product, market, progress, insight, business model, team and ask.

Open source ↗ · Evidence notes

Geoff Ralston

A Guide to Seed Fundraising

2016-01-07 · Guide

A practical introduction covering readiness, amount, dilution, financing forms, meetings, negotiations and closing.

Open source ↗ · Evidence notes

Tomasz Tunguz

The Harness Margin Opportunity

2026-09-17 · Investor analysis with illustrative economics

A model of how different orchestration costs can produce very different gross profit at the same contract price.

Open source ↗ · Evidence notes

Anton Osika

Anton Osika, Co-Founder and CEO @ Lovable: Hitting 85% Day 30 Retention - Better than ChatGPT

2025-03-05 · Founder interview and publisher takeaways

Discusses seed size, taking Series A money, retention and launches. Publisher takeaways explain why Lovable declined YC.

Open source ↗ · Evidence notes

02 / People and practice

Who to learn from

Fit judgments and limitations reflect the supplied research.
Person / fitWhy follow themMethods / where to begin
Anton Osika

Lovable co-founder and CEO

Strong first-person operator; capital allocation and traction positioning
Founder discussion makes dilution, focus and seed choices concrete; paired investor testimony explains underwriting.

Edited publisher takeaways for interview; no private financing documents or negotiation reconstruction.

  • Weigh financing against dilution and execution focus.
  • Use activation and retention evidence rather than headline user counts.
  • Explain an expanded market with concrete customer-built outcomes.
Start: Anton Osika, Co-Founder and CEO @ Lovable: Hitting 85% Day 30 Retention - Better than ChatGPT

Anton Osika, Co-Founder and CEO @ Lovable: Hitting 85% Day 30 Retention - Better than ChatGPT · Lovable Raises $200M, Valued at $1.8B, Just Eight Months After Launch · Accel’s $200M Series A in Lovable: Enabling the Last 99% · Lovable raises $330M to power the age of the builder · We just raised $400M in Series C funding to help people run their businesses

Michael Truell

Cursor/Anysphere co-founder and CEO

Strong product operator; fundraising process less directly evidenced
Useful for technical moat, professional-user positioning and capital tied to research capacity.

Interviews inspected chiefly through publisher descriptions; not a full account of investor selection or negotiation.

  • Choose a narrow initial customer and own a critical workflow surface.
  • Pair technical differentiation with dated adoption evidence.
  • Separate revenue definitions and research uses of funding.
Start: Michael Truell: How Cursor Builds at the Speed of AI

Series A and Magic · Series C and Scale · Past, Present, and Future · The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO) · Michael Truell: How Cursor Builds at the Speed of AI

Aravind Srinivas

Perplexity co-founder and CEO

Strong first-person fundraising-process operator
Primary Haas transcript describes how he handles actual investor questions and follow-up diligence.

His financing anecdotes are self-report; celebrity investors and prior research network limit direct comparability.

  • Use memo and live Q&A to interrogate the investment case.
  • Make metrics retrievable during diligence.
  • Ground category story in a distinct customer experience and shipping evidence.
Start: Perplexity AI CEO Aravind Srinivas, PhD 21, on why he ditched pitch decks

Perplexity AI CEO Aravind Srinivas, PhD 21, on why he ditched pitch decks · Perplexity AI Launches iOS Version of Conversational Answer Engine, Announces $25.6M Series A Round Led by NEA · Our Investment in Perplexity AI: Answer Engines and the End of Traditional Search · We are happy to announce that we’ve raised $73.6 million in Series B funding led by IVP

May Habib

Co-founder and CEO, WRITER

Strong first-person operator
Most directly applicable fundraising process, board selection and enterprise retention lessons in this lane.

Experienced repeat founder with established enterprise relationships; historical retention and ROI are self-reported, not current guarantees.

  • Use a raise to align the team on a three-to-six-year strategy.
  • Select board capabilities backwards from the desired company.
  • Prove renewal expansion through customer business outcomes.
Start: WRITER CEO May Habib on Full-Stack Enterprise AI

WRITER CEO May Habib on Full-Stack Enterprise AI · 4 things May Habib has learned as a founder, raising $100M, and building a board · WRITER raises $200M Series C at $1.9B valuation to fuel leadership in agentic enterprise AI · How Writer CEO May Habib Is Making GenAI Work For the Enterprise

Bret Taylor

Co-founder, Sierra; former Salesforce co-CEO

Strong first-person operator
Detailed enterprise pricing and selling guidance plus founder-authored capital-allocation announcements.

Celebrity founder with Salesforce, Facebook, Google and OpenAI ties; capital access and enterprise introductions are not replicable assumptions. Round figures are company disclosures.

  • Sell measurable completed jobs.
  • Study the customer's business before pitching.
  • Use design partners and production deployments to support platform expansion.
Start: LIVE: How AI is Reinventing Software Business Models ft. Bret Taylor of Sierra

LIVE: How AI is Reinventing Software Business Models ft. Bret Taylor of Sierra · There's an agent for that, and it runs on Sierra · Better customer experiences. Built on Sierra

Clay Bavor

Co-founder, Sierra; former Google Labs leader

Strong first-person operator; jointly authored sources
Co-authors Sierra's deployment and financing narrative with Taylor.

This lane has joint essays, not an individual Bavor fundraising-process interview; exceptional prior executive network.

  • Expand from support to customer-lifecycle outcomes.
  • Substantiate a platform claim with named deployments and implementation speed.
Start: Better customer experiences. Built on Sierra

There's an agent for that, and it runs on Sierra · Better customer experiences. Built on Sierra

Aidan Gomez

Co-founder and CEO, Cohere

Strong first-person operator; frontier-lab context
Explains private enterprise deployment, strategic distribution and sovereign infrastructure positioning.

Transformer-paper co-author, elite research network, strategic/government capital and frontier-model economics differ greatly from application startups.

  • Find security constraints that block production.
  • Start with a few high-value workflows.
  • Choose partners for compute and distribution.
  • Make national/strategic investor interests explicit.
Start: Inside the Paper That Changed AI Forever - Cohere CEO Aidan Gomez on 2025 Agents

Cohere raises $500M at $6.8B valuation to accelerate enterprise efficiency with agentic AI · The business case for generative AI: A conversation with Cohere's CEO · Inside the Paper That Changed AI Forever - Cohere CEO Aidan Gomez on 2025 Agents · Who Gets to Build the Frontier for AI · Cohere Announces Series E Fundraise Targeting CAD $7 Billion to Meet Surging Global Demand for Secure, Sovereign AI

Scott Wu

Co-founder and CEO, Cognition

Strong first-person operator; exceptional technical founder
Combines dated growth/burn disclosures, acquired distribution and concrete enterprise value measurement.

Elite programming reputation and investor access; software-agent demand is adjacent to the product, not proof of its market. Megaraises are not a reproducible process.

  • Separate organic growth from acquired ARR.
  • Explain product complementarity and cross-sell.
  • Measure useful completed work and commercial payback.
Start: Funding, growth, and the next frontier of AI coding agents

Funding, growth, and the next frontier of AI coding agents · Do it all with Devin: Announcing our Series E · The Wu Tapes: Q&A with Cognition's Scott Wu · AI should earn its keep: Introducing the AI Productivity Guarantee

Michael Seibel

Former YC accelerator leader; co-founder/former CEO of Justin.tv/Twitch and Socialcam; personal site says retired from YC in 2026

Investor/adviser with first-person founder experience
Best starting point for clear pitch language, specific evidence and separating fundraising from progress.

Core material dates from 2016–2019; not an active-investor directory. Current role checked on personal homepage.

  • Seven-question pitch and email comprehension test
  • Progress per elapsed time
  • Value-creation milestones
  • Concise investor-email preparation
Start: How to Pitch Your Company

How to Pitch Your Company · Fundraising Rounds Are Not Milestones · How To Cold Email Investors - Michael Seibel

Geoff Ralston

Seed investor/adviser; former YC president; authored YC seed guide in 2016

Investor/adviser
A compact end-to-end mechanics reference with a useful milestone-and-budget approach.

Historical US seed overview; sample monetary and dilution numbers are dated and instrument terms need current review.

  • Raise-to-milestone planning
  • Alternative round-size operating plans
  • Explicit meeting next steps and closing discipline
Start: A Guide to Seed Fundraising

A Guide to Seed Fundraising

Elad Gil

Entrepreneur, investor and author; co-founder of MixerLabs and Color Health; current site describes Gil&Co as a multi-stage investment firm

Investor/adviser with first-person operating experience
Strongest integrated sequence for capital need, process, partner selection, downside structure and AI defensibility.

Most financing craft dates from 2010–2015. Treat prescriptions about market structure and instruments as historical; AI moat essay is an argument, not a validated prediction.

  • Scenario-based capital and ownership planning
  • Synchronized investor process
  • Partner-level diligence and references
  • Honest milestone and cash updates
  • Build moats through customer workflow and execution
Start: 20 Questions To Ask Yourself Before Raising Money

20 Questions To Ask Yourself Before Raising Money · How to Raise a Successful VC Round or Series A · How To Choose The Right VC Partner For You · Financing approaches most likely to kill your company (if it is already on the edge) · Investor Update Emails · Defensibility & Competition · VC Insights on Investing in Artificial Intelligence with Sarah Guo and Elad Gil of No Priors

Sarah Guo

Founder of AI-focused venture firm Conviction; former Greylock investor

AI-specialist investor/adviser
Useful for how an AI investor separates novelty from usage, understands product reliability and evaluates technical/commercial ambition.

Investor thesis and personal judgment, not founder fundraising success proof. Full September 2026 Colossus transcript was gated; only public material accepted.

  • Probe conversion from waitlist to real usage
  • Separate model capability from a reliable product
  • Connect technical thesis to buyer and packaging
  • Understand investor objections without inventing certainty
Start: VC Insights on Investing in Artificial Intelligence with Sarah Guo and Elad Gil of No Priors

Launching With Conviction · VC Insights on Investing in Artificial Intelligence with Sarah Guo and Elad Gil of No Priors · No Priors, Just Conviction — Invest Like the Best, Episode 489

David Cahn

Growth investor at Sequoia; previously Coatue investor and venture leader, per official profile

AI-specialist investor/adviser
Explains what must connect infrastructure capital, economic value and a domain-specific application opportunity.

Mostly macro investor analysis and dated forecasts. Not a seed process handbook or evidence that the product fits Sequoia.

  • Make compute/revenue assumptions explicit
  • Differentiate by domain intent, data and workflow
  • Separate demand evidence from supply dependency
Start: AI’s $600B Question

AI’s $600B Question · AI in 2025: Building Blocks Firmly in Place · AI in 2026: A Tale of Two AIs

Tomasz Tunguz

Founder/general partner of Theory Ventures; previously Redpoint investor and Google product manager, per current personal biography

Specialist investor/adviser and quantitative analyst
Best complement for defending AI revenue quality, serving economics, retention and comparable-company limitations.

Analyses use external datasets and explicit illustrations; none should substitute for the company’s own cohorts and accounting.

  • Measure cost per successful outcome and gross-profit payback
  • Distinguish model switching from application retention
  • Normalize revenue definitions and valuation dates
  • Understand fund economics before sizing a round
Start: The Harness Margin Opportunity

Why Seed Rounds Are Growing as Startups Shrink · The AI Colander · The Harness Margin Opportunity · AI Harness’ ARR Multiples

03 / Put it to work

Practical playbooks

1. Build a case an investor can verify

A large ambition is more credible when the customer wedge, product evidence and expansion path fit together.

  1. Write a plain-language memo: customer, painful job, product, why now, differentiated insight, evidence, business model, team and capital needed.
  2. Show a live customer workflow and the underlying evidence. Separate actual usage/revenue from pipeline, waitlists, forecasts and reported run rate.
  3. Build a bottom-up market view from reachable buyers and plausible spending; label assumptions and show the initial wedge and expansion route.
  4. Use repeated questioning to expose weak assumptions. Update the memo and evidence, rather than adding unsupported claims to the pitch.

What to measure

Clarity without narrationEvidence coverageUnresolved objectionsCustomer references

Proposed preparation process. A concise memo complements the format investors request; a no-deck approach is not universally suitable.

How to Pitch Your Company · WRITER CEO May Habib on Full-Stack Enterprise AI · Perplexity AI CEO Aravind Srinivas, PhD 21, on why he ditched pitch decks

2. Prove demand, retention and AI economics

Show that customers keep getting value after the initial novelty and that delivering it can support a durable business.

  1. Define activation and retain cohort tables for actual target users. Keep paid accounts, active users, contracted ARR and annualized revenue distinct.
  2. Measure successful customer outcomes with the required quality level; include failed attempts, retries and human review.
  3. Record model/inference, infrastructure, support and human-service costs per workflow and account. Test the effect of heavier usage and changing model prices.
  4. Explain the source of durable differentiation: workflow integration, distribution, trust, proprietary access or feedback effects, with evidence for each.

What to measure

Account and revenue retentionCost per successful outcomeGross margin by cohortExpansionAcquisition payback

Use your product’s observed data. Model-provider retention, hypothetical margin examples and famous-company multiples are not the product benchmarks.

VC Insights on Investing in Artificial Intelligence with Sarah Guo and Elad Gil of No Priors · The Harness Margin Opportunity · AI Harness’ ARR Multiples · Defensibility & Competition

3. Size the raise around the next valuable milestone

Compare what different financing amounts allow you to prove, and what each costs in dilution and obligations.

  1. Define the next milestone that reduces a major business risk: retained paying demand, repeatable acquisition, enterprise deployment or another stage-appropriate proof.
  2. Model monthly operating cash with base and downside revenue, hiring, inference and infrastructure. Keep one-off financing and transaction costs in a separate schedule. Separate bookings from cash collection.
  3. A planning identity is: funding gap = expected net operating cash use through the milestone + contingency + one-off financing and transaction costs − deployable cash. Net operating cash use excludes the one-off costs added separately. Reconcile timing and committed funding before choosing an ask.
  4. Compare smaller, base and larger raises against speed, hiring execution, runway, ownership and future financing risk. Choose based on the operating plan, not peers’ headline rounds.

What to measure

Months to milestoneBase/downside cash runwayCapital per proof pointOwnership scenariosFinancing dependency

The formula is a planning identity, not a valuation or financing recommendation. Historical numerical heuristics are dated; actual terms and instruments require qualified transaction advice.

A Guide to Seed Fundraising · 20 Questions To Ask Yourself Before Raising Money · Financing approaches most likely to kill your company (if it is already on the edge) · Why Seed Rounds Are Growing as Startups Shrink

4. Select investors and manage a real process

Match the actual partner, stage, check size, ownership expectations and company needs.

  1. Build a research table of relevant partners using current public investment evidence. Record thesis, stage, geography, check/ownership assumptions, conflicts and why the partnership fits.
  2. Reference-check concrete help and behavior when a company struggles. Separate a fund’s marketing promise from founders’ experience.
  3. Prepare the core materials and practice before priority meetings. Run a bounded sequence with explicit owners, next steps and deadlines that allow real diligence.
  4. Track interest, diligence, written terms, executed documents and funds received as different states. Maintain honest alternatives; never invent competing offers.

What to measure

Qualified partner fitMeeting-to-next-step conversionTime in diligenceUnanswered diligence itemsCash received

Preparation only; no investors were contacted. Public portfolio evidence does not establish current mandate, willingness to invest or permission to share confidential data.

How To Choose The Right VC Partner For You · How to Raise a Successful VC Round or Series A · How To Cold Email Investors - Michael Seibel · A Guide to Seed Fundraising

5. Use AI as a fundraising analyst and preparation partner

Reduce preparation work while keeping factual claims and confidential disclosure under founder control.

  1. Organize an evidence room: company narrative, cap table, financial model, cohorts, customer references, product/security material and an assumption register.
  2. Have AI draft a claim-to-source map and challenging investor questions. Verify every answer against the actual source; keep missing evidence visible.
  3. Draft individualized meeting briefs and follow-up answers from an approved evidence set. Check numbers, confidentiality and attachments before any authorized sharing.
  4. Use objections and diligence questions to improve the product case, model and memo. Track repeated concerns and whether the underlying evidence resolves them.

What to measure

Preparation timeClaim traceabilityAnswer turnaroundCorrection rateRepeated objections resolved

Our proposed AI-assisted workflow. Srinivas’s AI-assisted preparation and live-Q&A account is self-reported; the complete workflow here was not tested on the product.

Perplexity AI CEO Aravind Srinivas, PhD 21, on why he ditched pitch decks · WRITER CEO May Habib on Full-Stack Enterprise AI · Investor Update Emails

6. Study large rounds without copying their advantages

Extract the investor logic and operating proof, then judge which parts transfer to your stage.

  1. For each selected case, read the founder explanation and the company/investor announcement together. Record the customer wedge, distribution, evidence and use of capital.
  2. Separate company traction from acquired revenue and annualized run rate; keep announcement dates aligned with the metrics they cite.
  3. Record founder reputation, prior exits, investor access, compute requirements and strategic partnerships that may not transfer.
  4. Write the smallest analogous proof your own product can produce. A large financing event is a case study, not proof that its publicly described tactic caused the round.

What to measure

Transferable mechanismsProof gapsStage-adjusted operating milestonesCapital efficiency

Selected success cases have survivor bias and unequal access. Announced terms and traction are primary-source claims, not independently audited financials.

Accel’s $200M Series A in Lovable: Enabling the Last 99% · Michael Truell: How Cursor Builds at the Speed of AI · Better customer experiences. Built on Sierra · Funding, growth, and the next frontier of AI coding agents

Capital / comparison

Funding rounds in context

Round amount, valuation and cumulative funding are distinct measures. Dollar figures from numeric fields are USD; other currencies and lower bounds are labelled explicitly. Missing figures are left explicit.
Company / foundersRound / announcedRound amount / currencyValuation / basisCumulative funding (USD)Lead investorsBusiness evidence / limits
Lovable

Anton Osika; Fabian Hedin

Series A

2025-07-17

announced as raised

$200M$1.8B

Unspecified pre/post-money

Not statedAccelInvestor cites 2.3 million active users and internal product use.

Exact closing date not disclosed; company and investor assertions not audited.

Lovable Raises $200M, Valued at $1.8B, Just Eight Months After Launch · Accel’s $200M Series A in Lovable: Enabling the Last 99%
Lovable

Anton Osika; Fabian Hedin

Series B

2025-12-18

announced as raised

$330M$6.6B

Unspecified pre/post-money

Not statedCapitalG; Menlo Ventures Anthology fundCompany cites project creation and traffic to customer-built apps.

Do not substitute traffic for revenue or project count for retention; no closing date disclosed.

Lovable raises $330M to power the age of the builder
Lovable

Anton Osika; Fabian Hedin

Series C

2026-08-12

announced as raised

$400M$13.3B

Unspecified pre/post-money

Not statedMenlo Ventures; Scaleup Europe Fund (managed by EQT; co-lead)Company cites more than 60 million projects and expanded business-workflow capabilities.

Round amount and valuation separately stated; cumulative funding not calculated; legal closing date unknown.

We just raised $400M in Series C funding to help people run their businesses
Anysphere / Cursor

Michael Truell

Series A

2024-08-22

announced as raised

$60MNot stated

Not disclosed

Not statedCompany reports more than 40,000 customers and specialized model/retrieval work.

Source names participating investors without explicit lead status. Founder list here names the profiled CEO, not every co-founder.

Series A and Magic
Anysphere / Cursor

Michael Truell

Series C

2025-06-06

announced as raised

$900M$9.9B

Unspecified pre/post-money

Not statedCompany reports over $500 million ARR.

Primary says $9.9 billion, not $9 billion seen in some databases. Lead and close date not stated; founder array not exhaustive.

Series C and Scale
Anysphere / Cursor

Michael Truell

Series D

2025-11-13

announced as raised

$2.3B$29.3B

Post-money explicitly stated

Not statedCompany reports over $1 billion annualized revenue, not explicitly ARR.

Primary names investors but does not label lead roles; founder array not exhaustive; not asserted to be latest corporate event as of research date.

Past, Present, and Future
Perplexity

Aravind Srinivas; Denis Yarats; Johnny Ho; Andy Konwinski

Series A

2023-03-28

announced as closed

$25.6MNot stated

Not disclosed

Not statedNEARelease reports 10 million monthly visits and two million unique visitors in February.

Closed by announcement; actual earlier closing day not given. Avoid conflating $25.6 million round with approximately $28.8 million offering or cumulative figures.

Perplexity AI Launches iOS Version of Conversational Answer Engine, Announces $25.6M Series A Round Led by NEA · Our Investment in Perplexity AI: Answer Engines and the End of Traditional Search
Perplexity

Aravind Srinivas; Denis Yarats; Johnny Ho; Andy Konwinski

Series B

2024-01-04

announced as raised

$73.6MNot stated

Not established by retrieved primary sources

Not statedIVPCompany reports ten million monthly users and over half a billion 2023 queries.

No primary-supported valuation, cumulative total or exact closing day entered.

We are happy to announce that we’ve raised $73.6 million in Series B funding led by IVP
WRITER

May Habib; Waseem AlShikh

Series C

2024-11-12

$200M$1.9B

Company states valuation; pre/post-money not specified.

Not statedPremji Invest; Radical Ventures; ICONIQ GrowthCompany cites named enterprise customers and average 9x ROI; investors cite full-stack differentiation and speed to value.

Company announcement, not audited traction; cumulative total not derived from incomplete history.

WRITER raises $200M Series C at $1.9B valuation to fuel leadership in agentic enterprise AI
Sierra

Bret Taylor; Clay Bavor

Additional financing; series label not specified

2025-09-04

$350M$10B

Company states valuation; pre/post-money not specified.

Not statedGreenoaksHundreds of customers and large-enterprise customer mix claimed; use of funds platform, US growth, Europe/Asia.

Founder/company claims and exceptional prior network; no audited revenue or margins.

There's an agent for that, and it runs on Sierra
Sierra

Bret Taylor; Clay Bavor

Financing; company says 'raising', series label not specified

2026-05-04

Announced as raising; close not independently verified

$950MMore than US$15B

Company says over $15B; exact valuation and pre/post basis undisclosed.

Not statedTiger Global; GVOver 40% of Fortune 50 served and named deployment timing and resolution outcomes, all self-reported.

Do not substitute secondary $15.8B precision for primary lower bound; announcement wording does not independently prove close.

Better customer experiences. Built on Sierra
Cohere

Aidan Gomez; Nicholas Frosst; Ivan Zhang

Funding round; series label not verified

2025-08-14

$500M$6.8B

Company states valuation; pre/post-money not specified.

Not statedRadical Ventures; Inovia CapitalEnterprise and sovereign AI expansion rationale; investor participation disclosed in primary indexed extract.

Full page body unavailable; round financials from indexed primary content.

Cohere raises $500M at $6.8B valuation to accelerate enterprise efficiency with agentic AI
Cohere

Aidan Gomez; Nicholas Frosst; Ivan Zhang

Series E initial close; targeting up to CAD $7B

2026-10-29

Initial close announced; total target is not closed proceeds

Target: up to CAD $7B; initial-close proceeds undisclosedCAD $35.7B if target achieved

Release states target would represent CAD $35.7B valuation; pre/post not stated; no FX conversion.

Not statedCompanies of Schwarz GroupCompany reports 165% year-over-year ARR growth and 300 customers; strategic compute, sovereign deployment and vertical-product investment.

Target is not closed proceeds. Exact initial-close amount undisclosed. USD fields intentionally null.

Who Gets to Build the Frontier for AI · Cohere Announces Series E Fundraise Targeting CAD $7 Billion to Meet Surging Global Demand for Secure, Sovereign AI
Cognition

Scott Wu; Steven Hao; Walden Yan

Funding round; series label not specified

2025-09-08

More than US$400M$10.2B

Explicit post-money.

Not statedFounders FundHistorical self-reported organic Devin ARR, net burn and low overlap with acquired Windsurf customers.

Primary says over $400M, so exact amount_usd null. Acquisition more than doubled ARR; do not represent all growth as organic.

Funding, growth, and the next frontier of AI coding agents
Cognition

Scott Wu; Steven Hao; Walden Yan

Series E

2026-09-08

More than US$2B$48B

Company states valuation; pre/post-money not specified.

Not statedAndreessen Horowitz; AccelCompany reports almost $900M revenue run rate and global enterprise deployments.

Primary says over $2B; exact amount_usd null. Run rate is not recognized annual revenue; no audited current burn.

Do it all with Devin: Announcing our Series E

Source catalogue

Find the next thing to read or watch

Starts with catalogued sources. Filter by person, topic, evidence or status; open a row for the method, claims and verification limits.

Anton Osika
2025-03-05
Anton Osika, Co-Founder and CEO @ Lovable: Hitting 85% Day 30 Retention - Better than ChatGPT

Discusses seed size, taking Series A money, retention and launches. Publisher takeaways explain why Lovable declined YC.

Founder interview and publisher takeaways
Catalogued source

Open original source ↗

20VC / Harry Stebbings · Capital versus focus and dilution

Method / workflow
Evaluate acceleration against dilution and distraction; competitor fundraising does not automatically require matching it. Track completed publishing as activation evidence.
Evidence
Primary firsthand
Verification / access
Original episode page and March 9 publisher takeaways read; transcript PDF retrieval failed; no full playback.
Source claims
Publisher attributes to Osika a preference for focusing on execution over copying competitors financing.; Publisher reports 85% month-one paying-customer retention; this is a cohort-specific founder claim.
Limitations
Takeaways are edited by the host; original full transcript not retrieved. Retention is not independently audited and cannot be compared across differing cohorts.
dilutioncapital-efficiencyactivationretention
Anton Osika
2025-07-17
Lovable Raises $200M, Valued at $1.8B, Just Eight Months After Launch

Announces a $200 million Series A at a $1.8 billion valuation.

Company financing announcement
Catalogued source

Open original source ↗

Lovable · Explain the market unlocked by the product

Method / workflow
Connect capital use to the nontechnical-builder market and a specific product ambition, rather than citing valuation as the thesis.
Evidence
Primary company/investor
Verification / access
Original publisher indexed article text inspected; exact date and financing terms visible.
Source claims
Company says it raised $200 million from Accel and other named investors.; Company positions the product for people who cannot code.
Limitations
Valuation pre/post-money basis and separate legal close date not stated.
roundmarket-positioninguse-of-funds
Anton Osika
2025-07-17
Accel’s $200M Series A in Lovable: Enabling the Last 99%

Accel explains market expansion, product use, founder judgment and the relationship preceding its investment.

Investor investment rationale
Catalogued source

Open original source ↗

Accel · Understand why an investor underwrites a company

Method / workflow
Show the investor actual product use and execution; develop conviction over repeated working interactions.
Evidence
Primary company/investor
Verification / access
Original investor article read.
Source claims
Accel says it led the Series A and used Lovable internally.; Accel describes more than ten trips between the teams countries over six months, and cites talent, execution and community.
Limitations
Investor advocacy, not an independent evaluation or evidence that meetings caused a financing.
investor-logicrelationship-buildingfounder-market-fit
Anton Osika
2025-12-18
Lovable raises $330M to power the age of the builder

Announces $330 million at a $6.6 billion valuation, led by CapitalG and Menlo Ventures Anthology fund.

Company financing announcement
Catalogued source

Open original source ↗

Lovable · Combine adoption evidence with a broader category

Method / workflow
Use customer-built outcomes and adoption evidence to explain the category and deployment of capital.
Evidence
Primary company/investor
Verification / access
Original dated article and financing/adoption sections read.
Source claims
Company reports more than 25 million projects in its first year.; Company reports 100,000-plus new projects daily.
Limitations
Projects and app visits are usage measures, not retained paid seats or realized revenue; valuation basis undisclosed.
roundtraction-qualitycategory
Anton Osika
2026-08-12
We just raised $400M in Series C funding to help people run their businesses

Announces $400 million at a $13.3 billion valuation, led by Menlo Ventures and co-led by the EQT-managed Scaleup Europe Fund.

Company financing announcement
Catalogued source

Open original source ↗

Lovable · Tie new capital to the next operating stage

Method / workflow
State what the next financing funds: business workflows, reliability, integrations, learning from customer outcomes and hiring.
Evidence
Primary company/investor
Verification / access
Original date, financing terms, product priorities and investor statements read.
Source claims
Company reports more than 60 million projects and 900 million monthly visits to built apps.; Company plans roughly 450 team members that year.
Limitations
Forward plans are not completed results; valuation basis and closing date not disclosed. Company usage claims are unaudited.
rounduse-of-fundsexpansion
Michael Truell
2024-08-22
Series A and Magic

Announces a $60 million Series A with product, customer and research milestones.

Company financing announcement
Catalogued source

Open original source ↗

Cursor / Anysphere Team · Make technical differentiation legible

Method / workflow
Pair the product vision with concrete technical capabilities and actual adoption.
Evidence
Primary company/investor
Verification / access
Original article read.
Source claims
Company reports more than 40,000 customers.; Company describes next-edit prediction, retrieval and speculative inference work.
Limitations
No valuation or named lead disclosed in this announcement; customer definition unspecified.
roundtechnical-differentiationadoption
Michael Truell
2025-06-06
Series C and Scale

Announces $900 million at a $9.9 billion valuation to advance AI coding research.

Company financing announcement
Catalogued source

Open original source ↗

Cursor / Anysphere Team · Connect scale to research investment

Method / workflow
Explain research capital needs using dated revenue and customer adoption evidence.
Evidence
Primary company/investor
Verification / access
Original date and article body read.
Source claims
Company reports over $500 million ARR and use at more than half of the Fortune 500.
Limitations
Use inside a Fortune 500 company does not establish an enterprise contract; ARR is company-reported. Valuation basis and lead not stated.
roundresearch-investmenttraction-quality
Michael Truell
2025-11-13
Past, Present, and Future

Announces $2.3 billion at a $29.3 billion post-money valuation.

Company financing announcement
Catalogued source

Open original source ↗

Cursor Team · Fund proprietary capability alongside distribution

Method / workflow
Show how adoption, research capacity and a durable product vision fit the next capital requirement.
Evidence
Primary company/investor
Verification / access
Original dated financing, revenue and use-of-funds text read.
Source claims
Company says annualized revenue crossed $1 billion.; Company says it has more than 300 staff and will invest deeply in research.
Limitations
Annualized revenue is not necessarily ARR; no separately verified accounting or legal close date. Investors named without lead designation.
roundpost-moneyrevenue-quality
Michael Truell
2025-05-01
The rise of Cursor: The $300M ARR AI tool that engineers can’t stop using | Michael Truell (co-founder and CEO)

Publisher chapters cover the IDE choice, product stack and defensibility.

Founder interview
Catalogued source

Open original source ↗

Lenny’s Podcast · Explain defensibility and the chosen customer

Method / workflow
Prepare a clear account of which user you serve, which interface you own and why model-level capability supports the experience.
Evidence
Primary firsthand
Verification / access
Original publisher-distributed Apple description and chapters inspected through indexed results; selected third-party transcript excerpts read; no playback.
Source claims
Publisher locates defensibility at 38:42 and the IDE decision at 18:31.
Limitations
Methods are study prompts derived from the topics, not proof of fundraising causation; transcript not read end to end.
positioningdefensibilityproduct-strategy
Michael Truell
2025-11-10
Michael Truell: How Cursor Builds at the Speed of AI

Publisher introduction emphasizes power users, owning the editor and work trials.

Founder and investor conversation
Catalogued source

Open original source ↗

a16z / Martin Casado · Choose a specific wedge instead of a generic AI story

Method / workflow
Articulate deliberate constraints and the customer wedge; defend why owning a workflow surface is necessary.
Evidence
Primary firsthand
Verification / access
Original dated publisher introduction read; audio not played and no full transcript obtained.
Source claims
Publisher says Truell explains choosing professional power users over a broad democratization narrative.
Limitations
Available evidence is publisher synopsis; specific fundraising mechanics not established. Host is an investor in the category.
wedgepositioningfounder-investor-dialogue
Aravind Srinivas
2025-10-21
Perplexity AI CEO Aravind Srinivas, PhD 21, on why he ditched pitch decks

Srinivas explains memo-led investor diligence and follow-up questions.

Founder talk with primary transcript
Catalogued source

Open original source ↗

UC Berkeley Haas · Run investor diligence through questions and evidence

Method / workflow
Prepare a memo, retrieve metrics live and answer the investor’s actual questions.
Evidence
Primary firsthand
Verification / access
Original date and relevant primary transcript sections read; no playback.
Source claims
He describes replacing later pitch decks with memos and live Q&A.; He recounts an AI-assisted follow-up preceding an investor transfer.
Limitations
Founder self-report; neither financing causality nor the private diligence exchange independently verified.
fundraising-processmemodiligenceAI-assisted-workflow
Aravind Srinivas
2023-03-28
Perplexity AI Launches iOS Version of Conversational Answer Engine, Announces $25.6M Series A Round Led by NEA

Company release explicitly says $25.6 million Series A closed, while introducing its iOS app.

Company press release
Catalogued source

Open original source ↗

Perplexity AI via PR Newswire · Pair financing with product and distribution milestones

Method / workflow
Show a live product, a distribution expansion and dated usage evidence alongside the funding story.
Evidence
Primary company/investor
Verification / access
Original release, issuer identity, date and closing wording read.
Source claims
Company reports 10 million monthly visits and two million unique visitors in February.; Funds support product and expansion work.
Limitations
Visits are not users; round amount is $25.6 million, not the approximately $28.8 million total/filing number shown in some databases. No valuation stated.
roundcloseddistributionmetrics
Aravind Srinivas
2023-03-28
Our Investment in Perplexity AI: Answer Engines and the End of Traditional Search

NEA states why the team, shipping pace, transparent answers and user experience attracted its investment.

Investor rationale
Catalogued source

Open original source ↗

NEA · Explain why the team can disrupt a difficult category

Method / workflow
Link founder-market fit to demonstrated shipping and a distinct user experience; show why this team can exploit a change in technology.
Evidence
Primary company/investor
Verification / access
Original dated investor article read.
Source claims
NEA explicitly cites founder-market fit and fast execution.; NEA confirms leading the Series A.
Limitations
Promotional investor testimony; not proof that pedigree or speed alone wins a term sheet.
investor-logicfounder-market-fittraction
Aravind Srinivas
2024-01-04
We are happy to announce that we’ve raised $73.6 million in Series B funding led by IVP

Company announces $73.6 million Series B led by IVP and names participants.

Official company financing announcement
Catalogued source

Open original source ↗

Perplexity · Support a financing claim with adoption evidence

Method / workflow
Keep the financing amount, investor roles and dated usage claims in separate evidence fields.
Evidence
Primary company/investor
Verification / access
Original company post body read; exact January 4 publication independently matched to the primary Perplexity blog index. Blog article direct retrieval failed.
Source claims
Company reports 10 million monthly users and more than half a billion queries in 2023.
Limitations
Valuation and cumulative total not established by the retrieved primary post; separate close date unavailable.
roundmetricssource-hygiene
Bret Taylor
Date unverified
LIVE: How AI is Reinventing Software Business Models ft. Bret Taylor of Sierra

Taylor explains Sierra's resolution-based pricing and the opportunity in domain-specific agents.

Interview with transcript
Catalogued source

Open original source ↗

Sequoia Capital · Value-based pricing and enterprise selling

Method / workflow
Research the customer's business before meetings; sell a valuable completed job; align charging with procurement and accountability.
Evidence
Primary firsthand
Verification / access
Original publisher's transcript inspected; no media playback. Publication date not displayed.
Source claims
Taylor says median customers pay an agreed rate for autonomous resolution; human escalations are free.
Limitations
Sequoia is an investor. Advice does not establish customer-level margins or make outcome pricing suitable for every contract.
pricingdistributionenterpriseinvestor-thesis
Bret Taylor; Clay Bavor
2025-09-04
There's an agent for that, and it runs on Sierra

Founders link a new round to platform development, US growth and international expansion.

Founder-authored announcement
Catalogued source

Open original source ↗

Sierra · Capital tied to platform and geographic expansion

Method / workflow
Connect new capital to specific distribution and product investments, supported by customer composition.
Evidence
Primary firsthand
Verification / access
Original company article inspected.
Source claims
Company announced $350M additional capital at a $10B valuation, led by Greenoaks.; Self-reported: hundreds of customers; over half had revenue above $1B.
Limitations
Company claims; no audited revenue, margins or fundraising process. Exceptional founders and existing investor network.
roundwhy-capitalenterprise-proofdistribution
Bret Taylor; Clay Bavor
2026-05-04
Better customer experiences. Built on Sierra

Sierra presents enterprise deployments and expansion across the customer lifecycle alongside a much larger financing.

Founder-authored announcement
Catalogued source

Open original source ↗

Sierra · Expansion from support into customer lifetime value

Method / workflow
Support a platform expansion story with named deployments and production timing, rather than market size alone.
Evidence
Primary firsthand
Verification / access
Original article inspected.
Source claims
Announced raising $950M led by Tiger Global and GV at a valuation over $15B.; Self-reported: serving over 40% of Fortune 50; Singtel resolution over 70%; Cigna authentication time down 80%.
Limitations
Uses 'raising', not an independently verified closing. Valuation is a lower-bound disclosure. Deployment metrics are self-reported.
roundenterprise-proofwhy-capitalexpansion
May Habib
2025-11-06
WRITER CEO May Habib on Full-Stack Enterprise AI

Habib uses fundraising to align the team on a multi-year story and stress-test convictions through repeated investor questioning.

Founder masterclass recap
Catalogued source

Open original source ↗

Radical Ventures · Fundraising as strategic alignment

Method / workflow
Move beyond quarterly plans; reconcile research, product and GTM into one long-term narrative, then test why each strategic choice exists.
Evidence
Primary firsthand
Verification / access
Investor-hosted session recap with direct founder statements inspected; no playback.
Source claims
Habib says continued payment does not itself prove continuing product-market fit.
Limitations
Investor-authored framing. No deck, terms negotiation, conversion funnel or complete process transcript.
fundraising-processstorystrategylimitations
May Habib
2024-05-28
4 things May Habib has learned as a founder, raising $100M, and building a board

Habib describes choosing directors backwards from the company she wanted to build and leading enterprise conversations with business impact.

Interview recap
Catalogued source

Open original source ↗

Insight Partners · Board composition and business-value positioning

Method / workflow
Choose board capabilities deliberately; map use cases by industry and learn from implementation.
Evidence
Primary firsthand
Verification / access
Publisher recap inspected; identifies Startup Grind original; video not played.
Source claims
Habib says board members were selected for varied perspectives and the company's future needs.
Limitations
Historical advice from an investor-hosted recap; a large board is not a universal recommendation.
boardinvestor-selectionpositioning
May Habib
2024-11-12
WRITER raises $200M Series C at $1.9B valuation to fuel leadership in agentic enterprise AI

Investors back a full-stack enterprise platform, blue-chip adoption and time to value; funding supports agentic workflows and vertical applications.

Official funding announcement
Catalogued source

Open original source ↗

WRITER · Investor thesis and capital use

Method / workflow
Present differentiated capability, production customer evidence and the product investments the money enables.
Evidence
Primary company/investor
Verification / access
Original announcement and investor quotations inspected.
Source claims
$200M Series C at $1.9B valuation, co-led by Premji Invest, Radical Ventures and ICONIQ Growth.; Self-reported average customer ROI of 9x.
Limitations
No valuation basis or audited ROI methodology disclosed. Do not generalize ROI across customers.
roundinvestor-thesiswhy-capitalenterprise-proof
May Habib
2024-03-05
How Writer CEO May Habib Is Making GenAI Work For the Enterprise

Habib explains a customer-outcome culture, regulated deployments and positioning beyond consumer productivity tools.

Interview with edited transcript
Catalogued source

Open original source ↗

Madrona · Retention, customer success and enterprise defensibility

Method / workflow
Make customer success an end-to-end business outcome role; document why the product remains useful when generic tools improve.
Evidence
Primary firsthand
Verification / access
Original transcript inspected; automatically generated and edited. Page says March 5; audio feed surfaced March 6; page date used.
Source claims
Self-reported historical NRR of 209%.; Habib describes six-figure mid-market deals and customer-managed cloud options.
Limitations
Historical metrics, not current performance. Product and model claims are not independent benchmarks; full-stack investment is capital-intensive.
economicsretentionmoatdistribution
Aidan Gomez
2025-08-14
Cohere raises $500M at $6.8B valuation to accelerate enterprise efficiency with agentic AI

Company ties funding to secure enterprise and sovereign AI and international growth.

Official funding announcement
Catalogued source

Open original source ↗

Cohere · Enterprise focus and global expansion

Method / workflow
Explain the buyer constraint and expansion plan that distinguish the company from general-purpose competitors.
Evidence
Primary company/investor
Verification / access
Original page opened but body was not exposed; title/date and indexed primary-page extract inspected.
Source claims
Announced $500M at $6.8B valuation in an oversubscribed round led by Radical Ventures and Inovia Capital.
Limitations
Limited body access; no independent financial verification or disclosed valuation basis.
roundwhy-capitalpositioning
Aidan Gomez
2024-02-23
The business case for generative AI: A conversation with Cohere's CEO

Oracle describes providing Cohere compute access and a global enterprise distribution platform, including Fusion applications and OCI services.

Partner interview
Catalogued source

Open original source ↗

Oracle · Distribution through a strategic platform partner

Method / workflow
Evaluate strategic capital and partnerships for buyer access, deployment infrastructure and integration—not only cash.
Evidence
Primary firsthand
Verification / access
Original partner-hosted article and founder statements inspected.
Source claims
Oracle states it integrated Cohere capabilities into Fusion applications and OCI's generative AI service.
Limitations
Oracle is a commercial partner and investor, not independent validation; compute economics and channel terms are undisclosed.
distributionstrategic-investorenterprise
Aidan Gomez
Date unverified
Inside the Paper That Changed AI Forever - Cohere CEO Aidan Gomez on 2025 Agents

Gomez explains that access to sensitive business context makes deployment control central to enterprise adoption; he recommends starting with the most important use cases.

Interview with transcript
Catalogued source

Open original source ↗

The MAD Podcast / Matt Turck · Private deployment as an adoption unlock

Method / workflow
Identify the trust constraint preventing production; prioritize a few high-value workflows before broad rollout.
Evidence
Primary firsthand
Verification / access
Original transcript inspected; no playback. Page date is 06/05/2025 with ambiguous numeric ordering, so ISO date left null.
Source claims
Gomez says VPC/on-prem deployment unlocks access to critical data.; He describes narrowing customer lists of hundreds of use cases to the top five.
Limitations
Founder perspective, not measured market-wide adoption. Research prestige and Google Brain network are unusual starting advantages.
enterprise-proofpositioningcustomer-discovery
Aidan Gomez
2026-10-29
Who Gets to Build the Frontier for AI

Cohere's founders frame financing around institutional control, trusted jurisdictions and a frontier alternative outside the US and China.

Founder-authored strategy essay
Catalogued source

Open original source ↗

Cohere · Sovereign AI as a capital and partnership thesis

Method / workflow
Connect product control and deployment requirements to the strategic interests of partners and buyers.
Evidence
Primary firsthand
Verification / access
Original founder byline, date and relevant article passages inspected.
Source claims
Announces initial close of Series E targeting up to CAD $7B, led by Schwarz Group with Government of Canada participation.
Limitations
Target is not closed proceeds; CAD is not USD. Government backing, frontier research credentials and national infrastructure aims are not a typical startup playbook.
roundinvestor-thesiswhy-capitalsovereignty
Aidan Gomez
2026-10-29
Cohere Announces Series E Fundraise Targeting CAD $7 Billion to Meet Surging Global Demand for Secure, Sovereign AI

Release details strategic investors, regulated-sector adoption and investment in compute, vertical products and global delivery.

Company-issued release
Catalogued source

Open original source ↗

Cohere via PRNewswire · Round status and infrastructure use of capital

Method / workflow
Separate signed/closed capital from target round size; show why capital requirements follow the operating plan.
Evidence
Primary company/investor
Verification / access
Company-issued release visible in detailed search extract; direct open returned inaccessible. Founder blog independently confirms target and initial-close status.
Source claims
Target CAD $7B; stated implied valuation CAD $35.7B.; Self-reported Q3 2026 ARR growth of 165% year over year and 300 customers.
Limitations
Release-only claims, not audited. No exact initial-close proceeds; no currency conversion applied.
roundenterprise-proofwhy-capital
Scott Wu
2025-09-08
Funding, growth, and the next frontier of AI coding agents

Wu separates Devin's pre-acquisition growth from Windsurf's added ARR and describes complementary products and a larger GTM channel.

Founder-authored funding announcement
Catalogued source

Open original source ↗

Cognition · Organic growth versus acquired distribution

Method / workflow
Show organic cohorts, burn and acquisition effects separately; explain how product adjacency creates cross-sell.
Evidence
Primary firsthand
Verification / access
Original company article inspected; proxy mirror excluded.
Source claims
Announced over $400M at $10.2B post-money, led by Founders Fund.; Self-reported: Devin ARR $1M to $73M by June 2025, historical net burn under $20M, under 5% enterprise customer overlap before Windsurf.
Limitations
All financial claims self-reported and historical; ARR is not recognized revenue. Acquisition creates comparability issues.
roundeconomicsdistributionenterprise-proof
Scott Wu
2026-09-08
Do it all with Devin: Announcing our Series E

Company pairs large financing with customer expansion, global offices and model-provider independence.

Official funding announcement
Catalogued source

Open original source ↗

Cognition · Capital for an independent agent lab and global delivery

Method / workflow
Tie funds to customer implementation capacity and technical independence; support the narrative with dated operating evidence.
Evidence
Primary company/investor
Verification / access
Original article inspected.
Source claims
Announced over $2B at $48B valuation, led by Andreessen Horowitz and Accel.; Self-reported run-rate revenue rose from $492M in May to almost $900M.
Limitations
Run-rate revenue is not annual recognized revenue. Exact proceeds and valuation basis unstated; no current margin or burn disclosure.
roundwhy-capitalenterprise-proofdistribution
Scott Wu
Date unverified
The Wu Tapes: Q&A with Cognition's Scott Wu

Long-form Q&A provides founder history and technical/product conviction beyond a funding headline.

Original founder Q&A
Catalogued source

Open original source ↗

Colossus · Founder conviction and early product limitations

Method / workflow
Use as a study of technical founder-market fit and willingness to ship and learn; do not treat exceptional talent mythology as a fundraising method.
Evidence
Primary firsthand
Verification / access
Original Q&A inspected; page gives May 2026 but no precise day, so date null.
Source claims
Publisher discloses Cognition is a sponsor and a Positive Sum portfolio company.
Limitations
Sponsor/investor conflict; profile mixes editorial claims and founder testimony. Technical celebrity and elite network limit transferability.
founder-market-fitlimitationscompany-building
Scott Wu
2026-06-04
AI should earn its keep: Introducing the AI Productivity Guarantee

Cognition estimates useful engineering output, excludes unmerged/unproductive work and supports an enterprise credit guarantee.

Company operating-method article
Catalogued source

Open original source ↗

Cognition · Auditable productivity and commercial accountability

Method / workflow
Prepare evidence of completed valuable work and specify measurement limits; distinguish productivity estimates from full business ROI.
Evidence
Primary company/investor
Verification / access
Original article's methodology and limitations inspected.
Source claims
Guarantee offers credits up to $10M if estimated engineering value falls below spend.; Company explicitly says its estimator does not replace ROI measurement.
Limitations
Credits are not a cash refund; human-equivalent effort is estimated. Commercial terms and measurement can affect applicability.
economicsenterprise-proofmeasurementlimitations
Michael Seibel
2016-07-19
How to Pitch Your Company

A seven-question pitch: product, market, progress, insight, business model, team and ask.

Article
Catalogued source

Open original source ↗

Y Combinator · Explain the company and make a specific ask

Method / workflow
Write a plain-language description; calculate a customer-specific bottom-up market; show progress relative to elapsed time; explain customer-derived insight. Have an uninformed friend restate the description without live clarification.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
Seibel prioritizes demonstrated progress and understanding over impressive language.
Limitations
Durable communication craft, not AI-specific underwriting or a guarantee of investment. Historical YC title in article is stale; his current personal homepage says he retired from YC in 2026.
timelessseedpitchmarket-sizingfounder-market-fitask
Michael Seibel
2017-02-06
Fundraising Rounds Are Not Milestones

A financing round is cash, not evidence of customer value or durable economics.

Essay
Catalogued source

Open original source ↗

Michael Seibel · Separate financing from business progress

Method / workflow
Define operating milestones around customer need, product-market fit, margins and sustainable value; assess the round as financing for that work.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
The essay rejects top-line growth with severely negative unit economics as a proxy for success.
Limitations
A principle rather than a numerical readiness rubric. Does not establish which milestones the product has reached.
timelessseedseries-amilestonesunit-economics
Michael Seibel
2019-09-24
How To Cold Email Investors - Michael Seibel

The official notes identify the information Seibel wants in a short investor email.

Video with official chapter notes
Catalogued source

Open original source ↗

Y Combinator · Prepare concise investor outreach

Method / workflow
Include problem, solution, launch status, growth, market, team and non-obvious insight in a message readable within a minute. Avoid origin-story padding, jargon, immediate in-person demands and rapid repeated follow-ups.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened official YC page; inspected publication date and timestamped chapter descriptions. Did not play full video.
Source claims
A deck is optional; if included, Seibel recommends a common format.
Limitations
Historical communication advice; not current permission to contact anyone. Email tracking is mentioned in the source but is not recommended here as a default.
timelessseedinvestor-targetingoutreach-preparation
Geoff Ralston
2016-01-07
A Guide to Seed Fundraising

A practical introduction covering readiness, amount, dilution, financing forms, meetings, negotiations and closing.

Guide
Catalogued source

Open original source ↗

Y Combinator · Size and execute a seed round

Method / workflow
Connect the ask to profitability or the next fundable milestone; compare operating plans for different amounts. Practice the pitch, establish explicit next steps, and distinguish verbal interest from signed documents and received funds.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date. Inspected amount, negotiation, closing and document sections.
Source claims
The guide frames capital as purchasing credible progress, with dilution as a tradeoff.
Limitations
Its 12–18 month horizon, round sizes, salaries and dilution ranges are 2016 heuristics, not October 2026 market benchmarks. US-oriented financing overview, not jurisdiction-specific legal guidance. Current SAFE forms were separately inspected at the official URL below.
timeless-frameworkdated-numbersseedraise-sizerunwaydilutionprocessterms
Elad Gil
2010-08-20
20 Questions To Ask Yourself Before Raising Money

A pre-raise decision checklist addressing purpose, ownership, financing alternatives and next milestones.

Checklist essay
Catalogued source

Open original source ↗

Elad Blog · Build a milestone-based capital plan

Method / workflow
Model cash costs and revenue under several scenarios; compare smaller and larger raises; specify the milestone each buys. Test founder ownership across future rounds, financing constraints and walk-away alternatives.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
Gil asks whether the next milestone enables breakeven or a stronger subsequent financing.
Limitations
The sample dollar amounts, 2–3x next-round valuation target and 50% padding are 2010 advice, not mandatory thresholds. Scenario modeling transfers; legal and instrument outcomes depend on actual documents.
timeless-frameworkdated-numbersseedgrowthraise-sizemilestonesdilutionalternatives
Elad Gil
2011-03-17
How to Raise a Successful VC Round or Series A

A tactical Series A process emphasizing timing, partner selection, iteration and negotiation alternatives.

Tactical article
Catalogued source

Open original source ↗

Elad Blog · Run a synchronized financing process

Method / workflow
Prepare and improve the pitch before priority meetings; compress first meetings into a bounded window, keep comparable options alive and know whether control, expertise or price matters most. Research the relevant individual partner.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
Gil argues that staggered meetings can leave founders comparing a live term sheet with investors who have barely started diligence.
Limitations
Written for 2011 VC structure. Claims about collusion, a small fixed number of firms and automatic doubled dilution in syndicates are not adopted as universal facts. Timing must allow real diligence; do not fabricate competing offers.
timeless-frameworkdated-marketseries-agrowthsequencingnegotiationinvestor-targeting
Elad Gil
2012-11-09
How To Choose The Right VC Partner For You

Evaluate the person joining the board as well as the fund brand.

Article and reference checklist
Catalogued source

Open original source ↗

Elad Blog · Choose an investor and board partner

Method / workflow
Assess relevant experience, chemistry, actual help, firm influence and board load. In references, ask for specific examples of useful contributions, bad-news reactions, weaknesses and behavior with other directors.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date. Inspected reference-check questions.
Source claims
The source distinguishes a partner’s practical ability and availability from firm reputation.
Limitations
Seniority generalizations reflect Gil’s 2012 experience; verify present partner capacity, mandate and conflicts individually. No current the product–investor fit established.
timelessseedseries-agrowthinvestor-targetinggovernancereferences
Elad Gil
2010-12-10
Financing approaches most likely to kill your company (if it is already on the edge)

An older warning about cap-table structure, follow-on signaling and a valuation the business cannot grow into.

Essay
Catalogued source

Open original source ↗

Elad Blog · Stress-test financing structure and the next round

Method / workflow
Run downside ownership scenarios before agreeing terms; examine how existing investors’ participation or absence may be interpreted in a later round. Read the actual down-round provisions.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
Gil argues that financing structure matters most when operating performance is mediocre rather than exceptional.
Limitations
Pre-SAFE 2010 context. Its categorical preference for notes without institutional VCs is not a current universal recommendation. Dilution, anti-dilution, governance and signaling depend on documents and circumstances; obtain qualified transaction advice.
dated-instrumentsseedseries-atermsdilutiondownsidesignaling
Elad Gil
2015-05-13
Investor Update Emails

Short seed and longer growth-stage investor-update formats.

Templates
Catalogued source

Open original source ↗

Elad Blog · Maintain a financing and assistance feedback loop

Method / workflow
Put specific asks first. Report highlights, lowlights and upcoming milestones; where appropriate add consistent metrics, cash and burn. Reuse a template and calendar the work.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
Gil recommends retaining investors’ practical involvement after the financing closes.
Limitations
Distribution is a founder decision: the source explicitly warns of later-stage leaks. This library authorizes no updates or outreach to investors.
timelessseedgrowthupdatesrunwayfeedback-loops
Elad Gil
2023-02-16
Defensibility & Competition

Gil challenges the idea that all useful young AI businesses must already possess a moat.

Essay
Catalogued source

Open original source ↗

Elad Blog · Explain how an AI wedge becomes defensible

Method / workflow
Separate current customer value from the mechanism that compounds defensibility: integrations, workflow breadth, distribution, contracts, platforms or truly differentiated data. Describe the path and evidence rather than claiming model access alone is unique.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
Many early SaaS and AI products are copyable; some build defensibility through subsequent execution.
Limitations
A taxonomy and investor opinion, not proof that every wrapper becomes durable. The source warns that data effects are often overstated. Specific 2023 platform-access examples are historical.
ai-specificseedseries-amoatwedgecompetitionmarket
Sarah Guo
Date unverified
Launching With Conviction

Guo sets out the rationale for an AI-native investment firm and the distinctive problems of its portfolio.

Founder/investor thesis
Catalogued source

Open original source ↗

Conviction · Understand an AI-native investor’s underwriting logic

Method / workflow
Make the technical and commercial thesis concrete: model failure handling, hallucinations, per-customer customization, and packaging work that previously required labor. Evaluate investor help through outcomes such as customers, talent, data and compute.
Evidence
Primary company/investor
Verification / access
2026-10-29: opened original Conviction article and inspected full extracted text. No publication date displayed; date is null rather than inferred from crawl time.
Source claims
The launch thesis combines traditional software-business discipline with new AI product and commercialization problems.
Limitations
Fund-launch intentions are not evidence that all promised services or thesis outcomes occurred. The historical first-fund amount is not a startup round-size comparison. No current investment availability inferred.
ai-specificseedinvestor-thesispackagingreliabilityinvestor-targeting
Sarah Guo; Elad Gil
2023-05-02
VC Insights on Investing in Artificial Intelligence with Sarah Guo and Elad Gil of No Priors

A first-person conversation about investor conviction, waitlists, product-market fit and whether a company needs VC.

Original interview with transcript
Catalogued source

Open original source ↗

The Cognitive Revolution · Distinguish AI excitement from usage and venture fit

Method / workflow
Test what happens when waitlisted users actually receive access. Show repeated usage rather than projecting signups. Compare venture financing with a cash-generating business, and model whether proprietary training is necessary for the specific product.
Evidence
Primary firsthand
Verification / access
2026-10-29: opened host’s original page and read timestamped transcript around 34:32–47:50, including waitlist, bootstrap and model-spending discussion. Full media not played; date displayed May 2, 2023.
Source claims
Guo questions waitlists that do not convert to real usage; Gil argues some companies should bootstrap and avoid unnecessary model-training spending.
Limitations
2023 market and cost examples are historical. Speaker opinions and anecdotes, not audited performance or present training-versus-API economics.
ai-specificseedrevenue-qualityusageretentionventure-fitcapital-efficiency
Sarah Guo
2026-09-01
No Priors, Just Conviction — Invest Like the Best, Episode 489

Guo discusses investing without historical backtests, investment decisions and raising her own fund.

Original interview; public excerpt and chapter guide
Catalogued source

Open original source ↗

Colossus / Invest Like the Best · Study investor judgment and honest fundraising narrative

Method / workflow
Study the investment-decision section at 22:15 and fundraise section at 28:15. Use objections to improve the case while distinguishing what is known, believed and still being tested.
Evidence
Primary firsthand
Verification / access
2026-10-29: inspected original host page, episode number, date, public introduction, transcript excerpt and timestamp list. Full transcript requires login; no login or full playback performed. The detailed fundraise story was only discovered in secondary transcripts and is not treated as verified in this record.
Source claims
The host describes her fundraise as grounded in execution rather than a polished story.
Limitations
Partial access. This is primarily an LP/fund-manager raise, so startup founders should not copy the process or assume equivalent reputation and access. September 1 date is supported by host and distributor, not crawl date.
current-2026ai-specificinvestor-judgmentnarrativepartial-access
David Cahn
2024-06-20
AI’s $600B Question

A dated model contrasting infrastructure investment with the revenue required to justify it.

Investor analysis
Catalogued source

Open original source ↗

Sequoia Capital · Connect compute spending to paid end-user value

Method / workflow
Expose assumptions linking compute purchase, total ownership cost and gross margin. In a founder pitch, explain who pays and why the application creates enough durable value to support its serving costs.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
Cahn distinguishes a transformative technology from the returns earned by investors financing infrastructure.
Limitations
The headline is a June 2024 scenario based on assumptions, not a measured present revenue shortfall, startup TAM or fundraising target. Infrastructure economics do not transfer directly to AI application margins.
dated-2024ai-specificunit-economicsgrowthcapital-intensityvalue-capture
David Cahn
2024-12-09
AI in 2025: Building Blocks Firmly in Place

Cahn describes professional AI search differentiation through intent, proprietary data, presentation and workflow placement.

Outlook essay
Catalogued source

Open original source ↗

Sequoia Capital · Make a domain-specific AI differentiation case

Method / workflow
Specify the buyer’s domain, required data, answer format and interface. Use those differences to explain why a specialized product earns adoption alongside general models.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
The essay argues that distinct professional workflows can support differentiated AI products.
Limitations
Predictions for 2025, including capex stabilization, are historical hypotheses rather than October 2026 facts. No independent forecast scorecard completed. Fundraising application is our synthesis.
dated-forecastai-specificwedgemarketpositioningcompetition
David Cahn
2025-12-03
AI in 2026: A Tale of Two AIs

Cahn forecasts continuing application adoption alongside possible compute, construction and AGI-timeline delays.

Outlook essay
Catalogued source

Open original source ↗

Sequoia Capital · Separate application adoption from infrastructure dependency

Method / workflow
Prepare separate evidence for customer demand and for the dependencies required to serve it. Stress-test runway and margins under delayed capacity or changing compute cost; do not let a macro AI story replace customer evidence.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
The essay treats application market pull and infrastructure supply constraints as different economic questions.
Limitations
A December 2025 outlook, not confirmed October 2026 outcomes. Revenue-per-employee examples and the predicted billion-dollar club are not readiness requirements. The practical stress test is our application of the thesis.
dated-forecastai-specificgrowthrunwaydependenciesmarket-pull
Tomasz Tunguz
2025-07-28
Why Seed Rounds Are Growing as Startups Shrink

An analysis of larger seed rounds despite smaller company teams.

Investor analysis
Catalogued source

Open original source ↗

Tomasz Tunguz · Understand fund economics without copying round size

Method / workflow
Ask how an investor’s fund size and ownership model fit the round actually needed. Compare capital required for the operating plan with the investor’s incentives before accepting a larger raise.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
Tunguz attributes much of seed-round expansion to investment-market structure rather than simply higher employee costs.
Limitations
The statistical decomposition is author analysis using outside datasets, not causal proof. 2025 market data is dated. Larger observed rounds are not a prescription for the product.
dated-2025seedfund-economicsraise-sizeinvestor-targeting
Tomasz Tunguz
2026-07-13
The AI Colander

Tunguz contrasts model usage retention with other software and observes rapidly shifting cost/performance leadership.

Investor analysis
Catalogued source

Open original source ↗

Tomasz Tunguz · Interrogate AI retention and switching

Method / workflow
For fundraising, separate new-user acquisition from cohort persistence. Measure the actual application’s account and revenue retention, and explain why customer workflow remains valuable when underlying models change.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date.
Source claims
The source argues that low model loyalty and changing price/performance increase buyer leverage.
Limitations
The cited retention relates to models and provider cohorts, not enterprise application logo retention or NRR. External studies were linked but not independently replicated. Do not transplant its percentages into an the product forecast.
current-2026ai-specificretentionrevenue-qualityswitchingcohorts
Tomasz Tunguz
2026-09-17
The Harness Margin Opportunity

A model of how different orchestration costs can produce very different gross profit at the same contract price.

Investor analysis with illustrative economics
Catalogued source

Open original source ↗

Tomasz Tunguz · Show unit economics per completed AI outcome

Method / workflow
Build an observed cost-per-successful-task ledger, including inference, hosting, evaluations and human checking; examine cost changes under heavier usage. Connect gross profit to customer acquisition payback rather than presenting revenue alone.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date. Inspected methodological footnotes distinguishing measured benchmark costs from hypothetical business assumptions.
Source claims
Tunguz uses an illustrative contract to show that serving cost changes margins and CAC recovery time.
Limitations
The companies, contract economics and margins are hypothetical. The cited coding-harness study is not proof of equal quality in other workflows; its sample cannot establish absence of quality differences. Reproduce cost/quality tests on the actual product.
current-2026ai-specificinference-costgross-margincac-paybackunit-economics
Tomasz Tunguz
2026-08-09
AI Harness’ ARR Multiples

An analysis of AI-company valuation dispersion that discloses substantial data limitations.

Investor valuation analysis
Catalogued source

Open original source ↗

Tomasz Tunguz · Normalize revenue and valuation comparisons

Method / workflow
Label annualized revenue versus contracted recurring revenue; align valuation and revenue dates; separate transaction-linked income from subscription economics. Explain category position and business quality instead of applying a famous company’s multiple.
Evidence
Primary advice/research
Verification / access
2026-10-29: opened original URL with web tool and inspected extracted article text and displayed publication date. Inspected footnotes on unaudited private data, estimated revenue and date mismatches.
Source claims
The source warns that some observations are estimates and that Ramp interchange revenue is not pure contracted subscription.
Limitations
Not a live comps service or a valuation formula. Private company data is unaudited; mismatched measurement dates can inflate multiples. None of the cited companies establishes a reproducible financing path for the product.
current-2026ai-specificgrowthrevenue-qualityvaluationcomparables

How to read the evidence

  • Research cutoff: October 29, 2026. This is bounded discovery of relevant public content, not every interview or a complete social timeline. Dates refer to publication unless a record explicitly explains otherwise.
  • No full recording playback or independent company financial audit was performed. Each record states whether evidence came from original text, show notes, indexed content, a recap or a secondary transcript.
  • Catalogued means the source is identified and useful, not that every claim is verified. Evidence labels and access limits must be read together. Discovery leads remain separate.
  • Shared interviews count once even when multiple people use them as a starting source. Individual source totals can overlap. Mirrored links are retained as related material.
  • Bounded current web verification, not exhaustive lifetime interview or financing history. Historical source dates preserved.
  • No full episode playback. Sources inspected only as described in each verification field.
  • Some original Perplexity blog and Lenny newsletter routes failed retrieval; official releases, company posts and publisher-distributed metadata were used where available.
  • Later Perplexity financing reports and secondary Cursor 2026 financing/acquisition tables were not promoted to confirmed rounds without inspected primary announcement evidence.
  • Company metrics and investor testimonials are attributed claims, not independent financial diligence. ARR, annualized revenue, visits, users and projects are distinct.
  • Round list is selected evidence, not a complete cap-table history; cumulative totals intentionally null.
  • Bounded primary-source scan, not a complete public corpus or a financing-history database.
  • 16 distinct records across five people and four companies; joint Sierra authors share two sources.
  • Text/transcripts and relevant passages inspected; no claim of full audio/video playback.
  • Sequoia and Colossus lack verified exact publication days; MAD's numeric date is ambiguous. These dates remain null.
  • Cohere August 2025 body not exposed; today's PRNewswire direct open failed. Access limits are recorded on each item.
  • Company performance, ROI, ARR and run-rate figures are self-reported unless explicitly described otherwise; no financial statements audited.
  • No current investor fit, availability, probability of funding or guaranteed replicability is inferred.
  • Survivorship, celebrity-founder, frontier-lab and strategic-government capital biases are substantial.
  • No outreach, account access, external mutation or recursive delegation performed.
  • Bounded web research, not a complete corpus of every relevant public appearance. Twenty distinct original sources across six practitioners; mirrors, excerpts and reposts deduplicated.
  • Source access and displayed publication dates were checked on 2026-10-29. One undated original has date null. Full audio/video playback was not performed; video notes and transcript access limits are recorded per item.
  • Primary advice is evidence of the practitioner’s method or thesis, not independent proof that it causes successful fundraising. Company/investor financial anecdotes and outside datasets were not audited or replicated.
  • Historical monetary amounts, runway horizons, dilution heuristics, investor-market assumptions and forecasts are explicitly separated from durable principles. Current access does not make an old recommendation current.
  • The three Cahn essays are separate dated perspectives, not a set of verified forecasts. No forecast-performance scorecard was conducted.
  • US/venture-oriented sources dominate. Grants, crowdfunding, bank debt, jurisdiction-specific securities law and tax treatment are not comprehensively covered.
  • Official SAFE page was inspected as a supplementary current reference; no CLI, signing tool, account connection, transaction or investor outreach was used.
  • No full legal/tax interpretation, investor outreach, publication, purchase or external mutation performed. Only the assigned JSON file was written.

Take the library with you

Download the full structured index, including sources outside the current filters. Original publisher links remain attached to each entry.