WillItSurvive
YC Summer 2026 · United States

Perceptron ML

Two-person pre-revenue YC startup pitching AI tools for law firms and a separate event-alert app.

Little protection left
Assessed 26 July 2026 · Evidence as of 26 July 2026
Not much stands between this business and its pressures.
On the recordSince 26 July 2026 · Checkable from Sept 10, 2026 · Every call on the record Checkpoint: YC Demo Day pitch to roughly 1,500 investors
The call

A two-person team with no revenue is running two pitches at once in a market where a few well-funded rivals already have the law firms and the users.

Perceptron ML is a very early-stage company built by two people working on two different ideas at once: one is AI tools for lawyers (competing with Harvey, Legora, and CoCounsel, which already have thousands of paying customers and billions in valuation), and the other is an alert system that watches public data and notifies users when important events happen. The legal AI space is already crowded with well-funded competitors who arrived first and keep getting better; the alert system is just a layer on top of what Claude, ChatGPT, and Gemini can already do for free or cheap, which means frontier AI models will likely absorb that capability themselves. The one thing that could matter is if the founders pick a single, narrow problem in one space instead of splitting focus, and then find a specific type of customer they can serve better than the big players can, but the company has not shown that path yet.

Why
01
AI pressure: Claude, ChatGPT, Gemini with agentic tool use and web search. Frontier models already scrape public data and send alerts on their own, the same job Notyfi is pitching. Could AI do the core job? Yes, likely.
02
Bigger players: Harvey AI, Legora, Thomson Reuters CoCounsel. Harvey ($11B), Legora ($5.6B, $100M ARR) and CoCounsel (1M+ users) already sell AI tools to the law firms Perceptron ML is targeting. Could a giant copy it? Yes, easily.
03
What protects it. One founder's NVIDIA background in production agents and formal verification, if aimed at a single narrow legal workflow.
What would change this call
Whether Perceptron ML presents a single focused product at YC Demo Day, Sept 10, 2026. Shows if founders resolved the dual legal-AI vs. Notyfi identity confusion.
Whether Notyfi is shut down or merged into Perceptron ML within two quarters. Would show founders narrowing focus instead of splitting effort across two products.
Whether Perceptron ML announces a named law firm pilot or paying customer before Demo Day. Would show first real traction against Harvey AI, Legora, and CoCounsel.
Key numbers
  • Raised $500K convertible note from Y Combinator (2026-07-01)
  • Harvey AI valued at about $11B (2026-03-01)
  • Thomson Reuters CoCounsel passed 1M users (2026-01-01)
Named risks
  • No structural moat; incumbents Harvey, Legora, CoCounsel can replicate either pitch with far more capital. Severity: high
  • Core mechanic is LLM plus scraper over public data, natively absorbed by frontier model agentic tool use. Severity: high
  • Dual, unsettled positioning (law-firm vertical vs horizontal Notyfi alerting) shown by inconsistent site copy. Severity: medium
  • Two-person team, pre-revenue, single $500K note, no verified customers as of mid-2026. Severity: medium
Rivals, as read in July 2026
RivalDirectionEvidence
Harvey AIlosing toValued ~$11B March 2026 vs Perceptron ML's $500K note and no revenue
Legoralosing to~$5.6B valuation and ~$100M ARR by April 2026, dwarfing pre-revenue Perceptron ML
Thomson Reuters CoCounsellosing toPassed 1M users in early 2026 while Perceptron ML has zero named customers
Docket AlarmcontestingOwned by vLex/Fastcase since 2018, already offers docket alerts Notyfi also targets
Questions people ask
Is Perceptron ML profitable in 2026?
No. Perceptron ML is pre-revenue with no disclosed customers or logos as of June 2026. It incorporated in Palo Alto on June 16, 2026, has a two-person team, and its only funding is a $500K YC convertible note dated July 1, 2026, per CB Insights.
Could AI replace Perceptron ML's product?
Its own core mechanic, an LLM plus scraper watching public data for alerts, is the same agentic tool-use workflow frontier models like Claude and Gemini are absorbing natively via MCP, meaning any well-resourced rival or even in-house teams could replicate it without needing Perceptron ML at all.
Who is Perceptron ML's biggest competitor?
In legal AI, Harvey AI (valued ~$11B, March 2026) and Legora (~$5.6B valuation, ~$100M ARR, April 2026) dwarf it; Thomson Reuters CoCounsel passed 1 million users in early 2026. In event alerts, Docket Alarm has run docket monitoring since 2018.
Who owns Perceptron ML?
Perceptron ML is privately, founder-owned by siblings Michael Marcotte and Peyton Marcotte, incorporated as a California stock corporation on June 16, 2026 in Palo Alto. Y Combinator is the sole listed investor via a $500K convertible note dated July 1, 2026, per CB Insights.
Has Perceptron ML raised any funding?
Yes, one round: a $500K convertible note dated July 1, 2026, with Y Combinator as the sole listed investor, per CB Insights data as of June 30, 2026. No priced valuation has been publicly disclosed.
What does Perceptron ML actually build?
The company presents two parallel pitches under the same two founders: a bespoke AI system for law firms (timekeeping, intake, discovery, drafting) branded Perceptron ML, and a separate horizontal event-alerting product called Notyfi that monitors public data like filings, RFPs, and lawsuits and triggers alerts within minutes.
Timeline
  • 2026-09-10YC Demo Day scheduled, forcing function for positioning clarity
  • 2026-07-01$500K convertible note closed with Y Combinator as sole investor
  • 2026-06-23Bizprofile records show CA incorporation filing active in Palo Alto
  • 2026-06-16Perceptron Ml Inc. incorporated as California stock corporation in Palo Alto
Revision history
26 July 2026
First assessed. Little protection left. The text above is fixed as of this date and is not edited afterwards.
In the same band
Sources