Crosby.

A law firm that moves at startup speed

It is 9:14 on a Tuesday. A growth-stage sales team in San Francisco just got a signed term sheet, drops a vendor MSA into a Slack channel, and goes back to its standup. By 10:12 the redlines are back - marked, benchmarked against market terms, and ready to send. No partner's assistant. No retainer call. No invoice with six-minute increments. That is Crosby at work, and it is the most boring-looking revolution in legal services you will see this year.

Crosby is a registered law firm. It carries malpractice insurance. It employs licensed lawyers. What it does not do is bill by the hour, and it does not make you wait weeks for a routine contract. The firm pairs proprietary AI agents with human attorneys to review NDAs, MSAs, and data processing agreements - the unglamorous paperwork that quietly decides how fast a company can actually close business.

Legal was built to be slow

Here is the inconvenient thing about traditional legal work: the business model rewards delay. When a firm bills by the hour, speed is a tax on its own revenue. For a company trying to grow, that incentive runs exactly backwards. The contract becomes the choke point - the place where a closed deal goes to sit and think about what it has done.

Ryan Daniels watched this from the inside. Before Crosby he was a lawyer at Cooley, one of the marquee firms for tech startups, and later served as general counsel at a fast-scaling company. He kept running into the same wall: deals stacked up, revenue waited, and the bottleneck was almost never the substance of the contract. It was the queue in front of it.

Most legal-AI tools answered this by handing software to the in-house lawyer and wishing them luck. Crosby's founders thought that was solving the wrong half of the problem. The bottleneck was not the lawyer's reading speed - it was the entire structure around them.

A lawyer and an engineer walk into legaltech

Daniels teamed up with John Sarihan, an early engineer at the fintech company Ramp, where he had a front-row seat to how regulatory and legal friction slows a hyper-scaling business. Their bet was contrarian: do not sell software to law firms. Become the law firm, and rebuild it from scratch with AI at the center.

That distinction matters more than it sounds. A software vendor can make a lawyer 20% faster. A firm designed AI-first can change the unit of work entirely - from billable hours to reviewed documents, from weeks to minutes, from a quote to a fixed price. The AI does the first pass. A licensed attorney verifies, negotiates, and signs off. The client gets a guarantee, not a guess.

Submit in Slack. Redlines before lunch.

The mechanics are deliberately unremarkable, which is the point. A team sends a contract or a legal question through Slack, email, or a CLM integration. Crosby's agents handle the routine analysis - clause extraction, risk flags, benchmarking against market terms. A human lawyer reviews the output, negotiates the parts that need judgment, and stands behind the result with malpractice insurance. Pricing is fixed per document, somewhere in the range of a few hundred to a thousand dollars, so the invoice never surprises anyone.

What you can actually do with it: clear the legal queue without hiring a legal team, get Big Law-quality review at a predictable price, and stop letting a routine NDA hold a signed deal hostage for three weeks.

The numbers do the arguing

It is easy to promise speed. It is harder to show a curve. In 283 days, the value of contracts Crosby negotiated jumped from $30 million to more than $1 billion. Revenue grew 400% in the months between the Series A and the Series B. The customer list reads like a tour of the companies everyone is watching: Cursor, Clay, Cartesia, Unify, Alloy, Overjet, Rogo, Gumloop, and Tishman Speyer among them.

The backers match the customers. The cap table runs through Sequoia, Lux Capital, Index Ventures, Bain Capital Ventures, and Elad Gil - and even Cooley, the firm Daniels left, put money in. There is a certain irony in a Big Law firm investing in the company built to route work around Big Law. Crosby seems comfortable with it.

Turn legal from brake to accelerator

Crosby is careful about one thing in particular: the AI does not practice law. It assists. Every review passes through a licensed attorney, and the firm carries the same professional liability as any traditional practice. That is not a hedge - it is the product. The guarantee is what lets a sales team forward a Crosby review to a counterparty without flinching.

The mission is to make legal the part of a deal that speeds it up. Not a smarter toolbar for lawyers, but a different shape for the whole service: routine contracts handled at the speed and price growing companies actually need, with humans where judgment matters and machines where it does not.

The billable hour is on notice

The Series B is pointed at the next, harder problems: counterparty response simulations, voice negotiation agents, and a client oversight platform so customers can see the work as it happens. If a firm can review a contract in 58 minutes, the obvious next question is whether it can negotiate one in real time. Crosby is betting yes.

There is a larger wager underneath. Legal is a $300 billion industry that still mostly sells time. If the unit of value shifts from hours to outcomes - from the meter to the result - the firms that priced their slowness as a feature have a problem. Crosby is not the only company circling this idea, but it is one of the few that became a law firm to do it.

Back to that Tuesday morning. The sales team in San Francisco never thought about its lawyers once. The MSA went out, the deal moved, the standup ended on time. The contract used to be the slowest thing in the building. Now it is the thing nobody notices - which, if you have ever waited six weeks on a redline, is the most radical outcome of all.