Crosby co-founder Ryan Daniels on building an AI-powered law firm where agents will eventually negotiate contracts

Summary

Crosby, an AI-powered law firm, emerged from stealth on June 17, 2025, announcing a $5.88 million seed round. The company was co-founded by Ryan Daniels, a lawyer-turned-operator, and John Sarah, an early Ramp employee. Their core thesis is that contracts are the connective tissue of all economic activity and that the $40 billion U.S. contract review market has gone essentially unchanged for 50 years.

Crosby is structured as an actual law firm, not a software vendor. Clients — currently including Cursor, Clay, and Unifi — submit contracts via Slack, and Crosby's hybrid AI-and-lawyer team returns a reviewed document with a median turnaround of 58 minutes. The firm has completed over 1,200 contracts since launch, building institutional knowledge of each client's preferences and risk tolerances. The target customer is the fastest-growing startup, where legal bottlenecks directly constrain revenue.

The longer-term roadmap is more ambitious. Daniels points to Stanford Law research showing that AI agents given the same negotiating parameters as law students consistently reach better outcomes for both parties in minutes. Crosby's goal is to map the roughly 50 main sticking points in a standard commercial agreement, model counterparty preferences from companies like Adobe or Microsoft, and simulate months of negotiation agentically in near real time. Daniels frames that as a few years out, but positions it as the company's defining objective.

On the Microsoft displacement risk, Daniels is largely unbothered. Crosby runs heavily on top of Word but sees itself as complementary rather than competitive, and notes that more contracts reviewed means more Word licenses purchased for its attorney team.

Atrium is the unavoidable historical comparison. Daniels has spoken with nearly all of Atrium's co-founders, and the consensus read is "right idea, wrong time" — the company burned capital on NLP that is now a solved problem. Atrium's wedge of cheap Series A legal work also proved difficult to expand since companies graduate to traditional big law for later rounds. Crosby's answer is to focus on recurring commercial agreements rather than one-time financing events, building a flywheel of preference data that compounds with each contract processed.

The business model remains service-packaged even as automation increases. Daniels is explicit that he intends to preserve the experience of working with a lawyer — responsive, conversational, frictionless — regardless of how much of the underlying work is eventually handled by agents.