Ryan Daniels spent years at the intersection of law and technology, first as a corporate attorney at Cooley, then as general counsel at A.Team, a talent platform for engineers. He watched contracts pile up. He watched lawyers charge by the hour to mark up the same clauses they had seen a hundred times. He watched startups sign deals they should not have because there was no one to read them carefully. When he founded Crosby in September 2024, the bet was simple: automate the first pass, let humans handle the judgment.
What he built is a hybrid. Eight AI agents, built on top of models from OpenAI, Anthropic, and Google Gemini, handle the initial contract review. When a startup sends a vendor agreement through Slack, where Crosby's system called Bailiff lives, the agents flag risky clauses, suggest redlines, and produce a markup. A human lawyer then reviews and signs off before anything goes back to the client. The median turnaround is 58 minutes, according to the company. The average contract costs $250 to $1,000 depending on page count, according to Forbes.
The model is working. Crosby has reviewed roughly 13,000 contracts for about 100 clients including Cursor, Clay, Cognition, and UnifyGTM, figures the company shared with Forbes. Sequoia Capital and Bain Capital Ventures have invested. Index Ventures and Lux Capital led a $60 million Series B that reportedly values the company at $400 million. Cursor, the AI code editor backed by Andreessen Horowitz, has used Crosby to process 2,000 contracts and says review time fell by roughly 50 percent. The company claims revenue has grown about 400 percent since October 2025, a figure it has not independently verified.
These numbers are real. The clients are real. The lawyers' names on the documents are real, which matters: a legal opinion carries liability, and liability requires a human being in the loop. That is also the bottleneck no one covering this space wants to talk about.
The hybrid model solves a real problem in legal ops. Startups routinely lack in-house counsel and cannot afford outside firms for routine contracts. The billable hour punishes speed. Crosby's model is flat fee, fast turnaround, a lawyer's signature at the end. It is genuinely differentiated from both the $500-an-hour law firm and the pure software tool that promises to replace lawyers entirely. The second category has a trust problem: when something goes wrong with a contract, you want to know a human being looked at it. Crosby has that human being.
But the human being is also the constraint. As volume grows, Crosby needs proportionally more lawyers. Eight AI agents can handle 1,000 contracts in three weeks, the company told Upstarts Media. To double that, you need more lawyers. Law firms scale linearly with headcount. Software scales exponentially. Robin AI, LawGeex, and a dozen other contract automation tools all face versions of this problem, but pure software plays at least have the option to invest in model improvements that reduce the human dependency over time. Crosby's margin structure, built on a per-contract flat fee with a human reviewer baked in, has a ceiling that a pure software competitor does not.
This is not an argument that Crosby is a bad bet. Index and Lux presumably ran the numbers and disagree with this analysis. It is an argument that the $400 million valuation is being priced as if the software will eventually dominate the cost structure, when the current structure requires the human to remain central to every contract that crosses the platform. Whether that ratio improves over time, whether the lawyers become reviewers of last resort rather than active signers, is the actual question the valuation depends on. The Forbes headline promised faster deals. Faster is real. But the bottleneck is still sitting in a chair somewhere in San Francisco, reviewing your redlines.
Crosby's pitch is not that AI will replace lawyers. It is that AI will make lawyers faster. That is a more honest claim, and a more limited one. The honest version deserves to be examined on its own terms: can the company hire and retain enough lawyers to keep turnaround times low as demand grows, and does the flat-fee model preserve margins at scale? Those are questions a startup with a $400 million valuation should be able to answer. The fact that the Forbes article does not ask them is the difference between a press release and a story.
