Stealth watchlist · Personal conviction

Three companies
worth the meeting.

Stealth-stage bets I've spent time with — each one attacking a different layer of the AI stack: the data layer under legal work (ALT), the specialization layer above frontier models (Noeri), and the interaction layer that sits on top of the OS (Tada). Concise memos below.

One-liner

The Westlaw replacement — built on data incumbents can't get.

What

An AI-native legal research platform. Rather than bolt a chatbot onto an existing corpus (Harvey/Legora do this — Harvey's US research literally proxies queries to LexisNexis), ALT is rebuilding the corpus itself. They're physically digitizing case law, statutes, and regulations — including microfiche/microfilm material that only exists in one or two institutions — and pairing it with a source-first search UX that sidesteps hallucination.

Why now

Legal research is a $4–5B US category with two incumbents (Westlaw, LexisNexis) charging $200–$1,000/user/month. Attorneys spend 19% of their working hours here. The bottleneck for challengers has always been data coverage — not model quality. ALT already has 85% of significant judicial opinions in key jurisdictions, and each newly digitized archive is a moat: the physical documents get harder to access every year.

Signal
  • Non-binding LOIs from 3 AmLaw 100 firms (>5,000 attorneys, >$5B in annual revenue)
  • Positive demo response from senior partners across the AmLaw 100
  • Data coverage already at 85% of Westlaw in target jurisdictions
  • Founder ran vendor evaluation at Wachtell — knows exactly what the buyer wants
Risks & open questions
  • Solo founder — technical hires and a strong #2 are the near-term unlock
  • Digitization is capex-heavy; capital efficiency depends on jurisdiction sequencing
  • Incumbents may respond with aggressive pricing or acquire an emerging leader (Casetext/vLex precedent)
My take

Legal tech has been won by whoever owns the data layer for four decades. AI raises the ceiling on what a better UI can do, but the floor is still coverage. ALT is the rare team credible on both sides — a Wachtell litigator with an AI background — going after the piece Harvey and Legora structurally can't build. 

Sources: founder-shared memos and decks, reviewed under NDA. Company codenames used where the team has not yet gone public. Views my own — not Menlo Ventures'.