Keye

What do they actually do

Keye builds an AI-first diligence platform for private equity teams. It connects to virtual data rooms, ingests messy PDFs and spreadsheets, and turns them into structured datasets that analysts can use. The software then runs standard investor analyses—like cohorts, retention, segment pivots, margin/driver decompositions—and links every number back to its source so teams can validate the math. Outputs can be exported as Excel workbooks with live formulas, not hardcoded cells, so they slot into existing workflows (Diligence Platform).

Users can iterate via a chat/command interface to ask for new cuts or follow-ups “like an analyst,” and the product flags anomalies and potential risks with source citations. The company positions itself for enterprise adoption with controls such as end‑to‑end encryption, SOC 2 Type II, and a stance of no data retention and no training on customer files (homepage). Keye launched publicly from stealth after private pilots and a seed round, and is now selling into PE firms as an enterprise product (PR Newswire, YC profile).

Who are their target customer(s)

• Deal analysts / associates: They spend hours extracting numbers from PDFs and rebuilding spreadsheets instead of testing investment hypotheses. The manual work is slow and error-prone, leaving little time for deeper analysis.
• VPs / partners running diligence: They need fast, trustworthy answers and consistent analyses, plus a clear audit trail to defend assumptions in IC. Today, tracing a figure back to a source document is tedious and unreliable.
• Operating partners / portfolio ops leads: They want standardized financial and customer metrics from diligence to run playbooks post-close. Today these metrics live in ad‑hoc files, making benchmarking and repeatable actions difficult.
• CIOs / security & procurement teams: They must ensure sensitive deal data is secure and compliant. Without strong assurances (no data retention, SOC 2, SSO/VDR integrations), large funds won’t approve firm‑wide deployment.
• Head of firm knowledge / deal memory owners: They need a single place to compare past diligences and surface prior findings. Fragmented, inconsistent files make it costly to reuse institutional knowledge or benchmark new deals quickly.

How would they acquire their first 10, 50, and 100 customers

• First 10: Convert private‑beta pilots and YC/seed intros by embedding a founder/PM in live diligence to deliver audit‑grade Excel outputs, closing short paid pilots in exchange for a case study and a named reference.
• First 50: Productize a “3‑deal” pilot (onboarding checklist, security pack, success metrics, fixed price), hire an AE with PE background plus SDR support, and run targeted outbound to mid‑market deal teams and ops partners, backed by early case studies and partner webinars.
• First 100: Shift to firm‑level enterprise sales with SOC2/compliance artifacts, SSO/VDR integrations, and standard contracts; staff enterprise AEs and CS; sell multi‑deal pilots that convert to firm licenses; add channels via VDRs, PE consultancies, and placement agents.

What is the rough total addressable market

Top-down context:

Global buyout deal count rebounded to roughly 3,000 in 2024, with deal value rising faster than volume—indicating a large and active diligence surface area for software to address (Bain 2025 outlook).

Bottom-up calculation:

If specialized diligence software like Keye is purchased on a per‑deal basis for buyout/growth transactions at an average of $25k per deal, then 3,000 deals implies ~$75M annual TAM for the diligence module alone; firm‑level licenses and portfolio‑ops add-ons would increase this.

Assumptions:

• Targetable deals are those requiring customer- and unit-economics analysis typical of buyout/growth diligences (~3,000/year worldwide) [Bain].
• Average per-deal software pricing of ~$25k for ingestion, structured analyses, risk flags, and Excel exports.
• Estimate excludes broader firm licenses and ongoing portfolio monitoring, which could materially expand TAM.

Who are some of their notable competitors

• Datasite: A leading virtual data room provider increasingly layering AI features for diligence workflows; relevant as both a distribution partner and potential competitor if analytics deepen.
• Luminance: AI contract review used in M&A and legal diligence; overlaps on document analysis and risk flagging across large corpuses of deal documents.
• AlphaSense: Research and document intelligence used by PE for market and competitive diligence; not purpose-built for VDR math, but competes for analyst time and insight generation.
• DataSnipper: Excel-native document and table extraction popular in audit; PE analysts sometimes repurpose it to speed PDF-to-Excel workflows—an incumbent manual alternative to Keye’s automation.
• Tegus: Expert calls and research library widely used in PE diligence; complements Keye but competes as a spend category within diligence tooling.