Blueshoe wants to be the law firm that never sleeps
A Harvard Law graduate and an MIT engineer are rebuilding the firm from the reasoning up - pairing licensed attorneys with AI that argues in caselaw, not summaries. The pitch fits on a bumper sticker: justice is leverage, not luxury.
Ask a normal person what happens when they have a legal problem and the honest answer is usually nothing. They read the letter twice, Google a phrase, and decide the hassle of finding a lawyer costs more than the problem itself. That gap - between having a right and being able to use it - is the thing Blueshoe decided to build a company around.
Blueshoe (YC X25) is a small firm with a large ambition: to be a law firm that runs on AI. Not a software vendor selling tools to other firms, and not a chatbot pretending to be a lawyer, but an actual firm where licensed attorneys take cases through resolution using reasoning software the company builds itself. It started as an AI platform for legal research and grew into the firm around it.
The company was founded in 2025 by Casey O'Grady and Kai Yee Wan, went through Y Combinator's 2025 batch, and has reported roughly $500K in seed funding. Two people. The problem they picked is not small.
The naming tells you something about how the company thinks. It launched publicly as "the AI platform for legal reasoning" and now describes itself as "the first tech-powered law firm for individuals and small businesses." That shift - from platform to firm - was not a rebrand for its own sake. It was Blueshoe following the problem downhill. The research tool was useful, but the people who most needed it were also the people least able to hire a lawyer at all. So the product grew a firm around itself.
01 - The problemThe efficiency the industry admits it doesn't have
Law has a strange relationship with its own inefficiency. Lawyers know their work is slow, and they mostly bill by the hour, which quietly rewards the slowness. Blueshoe likes to point at a survey figure: most lawyers say efficiency is a priority, yet only a fraction believe their own practice actually delivers it. The wish and the reality sit far apart.
For the person on the other side of that inefficiency, it shows up as price. Good representation is expensive, opaque, and hard to schedule. Blueshoe's framing - "justice is leverage, not luxury" - is aimed squarely at that experience. If the cost to serve a case can be pushed down with software, the argument goes, then the same quality of counsel can reach people who were priced out.
02 - The productSoftware that reasons, not just retrieves
The technical bet underneath Blueshoe is a claim about what legal AI should do. Most tools summarize. You feed them documents and they hand back a shorter version. Blueshoe's platform is built to go one step further and construct what the company calls reasoning chains - connecting facts to statutes and caselaw into an argument you can follow and audit back to the primary source.
The design goal doubles as a safety goal. O'Grady describes the product as built to prevent the familiar failure modes of legal AI: "hallucinations and bad reasoning, and lazy lawyering." A summary can be confidently wrong and no one notices. A reasoning chain that has to cite its way back to a source is harder to fake. It also runs on law-firm-grade security, with enhanced encryption and a stated policy of not training its models on user data.
03 - The customersFrom law-school labs to real clients
Blueshoe's first users were students. The platform runs pilots at Harvard, Yale, and Columbia law schools, where students, researchers, and faculty can register with a school email and use it to build and test arguments. It is an unusually credentialed testing ground for a two-person company, and a smart one - law students are demanding users who will find the weak spots fast.
On the firm side, the customers are individuals and small businesses - the segment that expects everything to be always-on, instant, and available everywhere, and finds the traditional legal industry to be none of those. Blueshoe takes consumer and plaintiff-side matters: employment, property and housing, personal injury, consumer protection, product liability, and commercial disputes. Guidance comes through a web app or an AI interface; the actual representation comes from attorneys.
not luxury."
04 - The differenceOwning the firm instead of selling to it
Most legal AI startups sell software into existing firms and hope the firms change. Blueshoe made a different structural choice: become the firm. That decision quietly rewires the incentives. When you sell tools by the seat, your customer is a partner protecting billable hours. When you are the firm, every efficiency you invent flows to your own margins and your own clients. The company can drop billable hours because it is not asking anyone else to.
| Approach | Legacy research | AI tool vendors | Blueshoe |
|---|---|---|---|
| Output | Search results | Summaries | Reasoning chains |
| Who they sell to | Firms | Firms | End clients |
| Pricing | Subscription | Per seat | Flat, no billable hours |
| Represents you | No | No | Yes, with attorneys |
That also frames who Blueshoe is really up against. In research, the incumbents are Westlaw and LexisNexis. In AI tooling, names like Harvey and CoCounsel. But as a consumer firm, the competition is the whole tangle of traditional plaintiff practices and the intake-and-forms services people currently settle for. Blueshoe is picking a fight on two fronts, which is either overreach or the only way the model works.
05 - The peopleTwo halves of one problem
The founding team splits the problem cleanly down the middle. Casey O'Grady, the CEO, took the long route into law - Rice, then finance, then Harvard Law, then strategy consulting where he advised large companies on adopting AI. He has seen both how the technology gets sold and how the profession actually works. Kai Yee Wan, the CTO, comes from the engineering side: MIT, software work at Google, and a stint teaching computer science.
It is a familiar shape for a hard startup - one founder who understands the domain deeply enough to know what is broken, one who can build the machine to fix it. In a field as unforgiving of hand-waving as law, having a real lawyer in the room is less a nice-to-have than a requirement. It also shapes the product's tone. Blueshoe is careful to describe its AI as iterative with lawyers rather than a substitute for them, which is both a legal-risk posture and a genuine belief about where the technology is useful today.
That belief has drawn the company into rooms where legal technology usually gets argued about rather than sold. O'Grady has taken the pitch to Harvard Law itself for a legal-tech talk, the kind of appearance that signals Blueshoe wants to be taken seriously by the profession, not just by consumers looking for a cheaper option. For an early-stage firm, credibility inside the academy is a form of distribution.
06 - The marketA big, slow category waiting for a shove
Legal services is enormous, essential, and famously resistant to change. Every prior wave of legal technology - from Westlaw digitizing the library to e-discovery to the current AI moment - has moved the industry a notch without breaking its basic economics. Blueshoe's wager is that AI is finally good enough to change not just the tools but the shape of the firm itself, and that being early to build that firm beats waiting to sell into the old ones.
It is early. The team is two people, the funding is a seed, and the leap from law-school pilots to a scaled consumer firm is exactly where legal-tech ambitions have stalled before. What Blueshoe has is a clear thesis, a credentialed proving ground, and a business model that lines its own incentives up with its customers'. Whether that adds up to the law firm that never sleeps is the thing the next few years will decide.