For VC partners and for founders raising
Drop a PDF. In five to seven minutes, get the memo a senior partner would write. Two verdicts, sourced TAM, named comparables, a risk read calibrated to the fund. Forwardable. Print-ready.
Real deck · identifier withheld
AI camera monitoring for seniors. Fall detection + behavioral activity for home care.
How it works
Benchmark suite
Caught Theranos, FTX, Fyre, WeWork, Zenefits cold. Cleared DoorDash, Coinbase, Shopify, Brex, Buffer. Every prompt change re-runs the suite before it ships.
Theranos · FTX · Fyre · WeWork · Zenefits · DoorDash · Coinbase · Shopify · Brex · Buffer
Compare to Bessemer
Pick a deck Bessemer evaluated. Verdict generates the memo from public source materials only. No insider info. No hindsight. Then click What Bessemer actually didto compare against the partnership’s real verdict.
Twilio-of-identity wedge with proven PMF: 75 paying customers, $10K MRR, 90%+ gross margin, ~0% net churn at v1. Developer-first IAM is the right shape for the moment.
Sits at the intersection of two BVP roadmaps (developer API + cyber security). Stage and check fit for a $1M seed extension at the existing $8M pre-money cap.
BVP's own analysis (page 11) acknowledges the SSO market is commoditizing. Okta and OneLogin sit at $30M and $10M ARR respectively, both targeting IT-led purchases. Auth0's wedge is the developer purchase. The CIO is downstream. Identity-as-a-service for the SaaS vendor side is structurally pre-empted by SSO incumbents only if those vendors solve the same multi-identity-source problem with the same documentation quality. They do not.
BVP's own memo (page 12) names DIY as the strongest competition for the SaaS-vendor segment. At $50K+/year, a vendor might roll their own. Vulnerability-of-the-week events (Heartbleed-class) raise the true cost of DIY beyond raw engineering hours.
As SaaS vendors mature, complex enterprise IAM needs may push them off Auth0 onto a heavier solution like ForgeRock ($30M Series C). Mitigation depends on whether Auth0 builds enterprise capability fast enough.
Stormpath and Gigya target adjacent segments (web-first, marketing-first); neither has a dedicated developer-first enterprise IAM play. Customers cited documentation as the reason they chose Auth0. That moat compounds with engagement.
Auth0 is what every developer roadmap firm wants to back: technically sharp founders (Eugenio Pace and Matias Woloski wrote Microsoft's identity book; Jon Gelsey joined as CEO from Microsoft + Intel Capital), 75 paying customers in 16 months, and ~0% net churn at the v1 product. The Twilio-of-identity framing is the right wedge into a category that's commoditizing on the IT side.
The $1M check at the existing $8M pre-money cap is the kind of opportunistic seed extension that maps directly to BVP's developer + cyber roadmaps. Super-pro-rata for 33% of the Series A is the right ball-control for a category leader. The risk register is real: DIY at the high end, graduate-out to ForgeRock at the enterprise end. Neither is a dealbreaker at the seed stage with this team.
That’s the artifact. Drop your deck and get the same memo.
Try Verdict freeFounder raising? Run your own deck cold, stress-test my deck →
The contrast
Built for analysts seeing fifty-plus decks a month. Tab one is the workflow many VCs default to today (drop a PDF into a chat model and see what comes out) lined up against the Verdict path. Tab two is the call Verdict made cold on three real decks where the public outcome already settled the question.
Both flows start with the same deck. One ends in a chat reply. The other ends in a print-ready memo a partner can forward to an LP.
The default path
Generic project workspace
The Verdict path
Diligence-grade output, partner-forwardable
Step 01
Upload deck PDF to Claude project
Drop the deck into a project folder. No extraction layer, no per-page indexing. It sits in context as a single attachment.
Step 01
Upload deck PDF to Verdict
Per-page text extraction runs first. Every later citation references a specific page, not the document as a whole.
Step 02
Paste a prompt asking for a VC memo
Something like 'write me a VC memo on this deck.' The shape of the output depends entirely on how the prompt is worded that day.
Step 02
Paste website plus founder LinkedIn for evidence
Optional, but doing it lifts the rigor of every claim. The chain pulls in real signals (recent press, founder track record, sector traction) instead of inferring.
Step 03
Wait roughly 2 to 3 minutes for Sonnet to draft
Single-pass generation. The model reads the deck and writes prose in one sweep, no retrieval, no follow-up calls, no structured rationale checks.
Step 03
Five-call analyst chain runs in five to seven minutes
Frame, market, risks, evidence, synthesis. Each call writes structured JSON. Each verdict carries a confidence band and a 'would change mind' clause.
Step 04
Read prose memo without fixed sections
The structure shifts run to run. Some drafts have a verdict line, some don't. Some cite the deck, some don't. Comparing two memos side-by-side is manual work.
Step 04
Read structured report with fixed sections
Verdict, fund-fit, market sizing, risk triangulation, GP summary. Same shape every run. Two memos can be diffed line by line.
Step 05
Output is a chat reply meant for copy-paste
No PDF. No share link. No Chicago footnotes. The artifact lives inside a chat thread.
Step 05
Output is a print-ready PDF and a share link
Chicago-style footnotes with source URLs. A partner can forward the link to an LP or save the PDF to a deal folder.
Output
A prose draft inside a chat window.
Output
A structured memo a partner can forward.
Frequently asked
Direct answers to the questions cold visitors ask. If something here is unclear, write back and it gets rewritten.
Decks live in isolated Supabase Storage scoped to the user account, get sent to Anthropic for chain processing under Anthropic’s API terms (which prohibit training on customer data), and are deleted whenever the user deletes the report. Reports auto-purge 90 days after creation if not deleted sooner. Sub-processors are Anthropic, Supabase, Vercel, Clerk, and Stripe.
SOC 2 status: not yet certified. SOC 2 Type I is on the Q3 2026 roadmap if traction warrants the audit cost. Procurement teams that need an attorney-reviewed data-handling addendum can request one through the in-app support form.