How This 25-Year-Old Built A $675M Legal AI Startup (With No Legal Experience)

AI is moving super, super quickly, and we have to match that. As we go deeper into the legal stack, the line between software and service blurs. Our strength has been saying, we do not know exactly where this goes—neither do you—so let’s partner to make sure we both win.

I am joined by Max Junestrand, CEO and co-founder of LaGora, the AI workspace for lawyers from YC Winter twenty-four. It has been thirteen months since the batch. What is LaGora, how did the ChatGPT moment shape it, and what did legacy tools look like? Also, give me the first time it felt magical for a customer.

We are the AI-powered workspace for lawyers—review, drafting, research—replacing a fragmented set of point solutions. GPT three point five was the unlock: we built a quick POC and then an enterprise system used by tens of thousands. Early on we focused on Europe—data residency, no retention, exemptions from human review—so firms could actually deploy it. The first magic: at Mannheimer Swartling we answered a complex research query perfectly by tying legislation into RAG; then our due diligence grid let firms ask hundreds of questions across hundreds of documents with instant, cited answers—work that took days now takes minutes.

You are announcing a Series B. How much, who led—and what does a lawyer’s day inside LaGora look like?

Eighty million dollars led by Iconiq and General Catalyst, with YC, Benchmark, and Redpoint. Day to day: a web app chat that has become an agent using tools via MCP to run step-by-step workflows—research, conform to firm language, produce the memo—plus Tabular Review that runs at massive scale with chunking and cross-references, and a Microsoft Word add-in—cursor for lawyers—that reads and edits documents and executes playbooks or multi-step negotiations in the right-hand pane.

Name something that was impossible before, and tell me why selling to law firms flipped from impossible to inevitable. Can LaGora negotiate on my behalf?

Old ML could not understand varied legal language; now we do semantic clause detection, automatic redlining against playbooks, and deep research that fuses judgments, legislation, and the web. Due diligence is becoming a commodity—clients expect it and will not pay for simple review—so firms must adopt AI. It is not a race to the bottom; AI frees time for high-stakes advice. People even role-play opposing counsel in our chat during hearings—it feels like extra armor. For negotiation, pure LLMs are not enough, so we use Playbooks: firm rules, approved language, and fallbacks. Press play and it marks up to standard. It scales beyond legal to sales, compliance, and risk, and enforces consistency.

You were not lawyers. How did you learn fast, and what should founders without domain expertise do?

We stayed humble, iterated daily with early partners, and hired lawyers. I interviewed roughly one hundred lawyers by inviting them to lunch and offering to pay their hourly rate. Be someone people want to help: ask a ton of questions and give ideas back. That got us up the curve quickly.

How do you outpace incumbents, what is under the hood, and who buys? Also, how do you wedge into a firm?

AI reset the category. With about one hundred people we out-ship teams many times our size; buyers now avoid five-year lock-ins and judge rate of change. We run on Azure, hot-swap models, and route simple queries to smaller models and complex ones to larger to keep margins tight. At big firms, innovation teams drive adoption with practice practitioners; at mid-size firms, partners decide. You cannot do bottom-up because of procurement and security, so start with one partner team, make them rockstars, then expand.

Give me your backstory, the growth from ten to one hundred, and how you changed as a leader and built culture.

At eighteen I chose college over going pro in Dota two, did dual engineering and business during COVID, coded eSports betting, spent time at Norrsken, McKinsey, and even a week at Depict. Post-YC we paused selling for about four to five months to harden onboarding and reliability—first experience matters—then opened hubs in Stockholm, London, New York, and placed people locally across Europe. I shifted from IC to delegating and hiring founders, seeding new hubs with our best Stockholm people. We screen for owners who just get things done. Flat org, AI-powered generalists—five marketers can do what used to take thirty—and I always ask candidates what they have done outside their role.

Fast forward five to ten years: what is a lawyer’s day? Are labs going vertical? How does product market fit feel? And why stay in Stockholm instead of moving to San Francisco—what is next?

Lawyers will instruct and review AI agents’ work and manage client delivery. Labs are becoming platforms; whatever they ship is table stakes, and we add the layers. Product market fit feels like drag—near-infinite demand and true reliance. We stayed in Stockholm to dominate the Nordics, expand across Europe, then hit the US as a shark, not a small fish—we opened New York and launched with a top firm. As AI blurs software and service, the category leader must be a strategic partner, so we scaled headcount while keeping urgency and culture.

Parting advice for vertical AI founders—and for candidates considering LaGora?

Do not lock into one lab or try to outrun them. Build moats that rise with model improvements; pick a narrow wedge or use models creatively—like scribing—to match domain language. For candidates: bring ambition and hunger—this is not nine to five. Expect case interviews: pitch our product if you are go-to-market, or build a LaGora POC if you are engineering. We reference deeply and hire across Europe—we are building a lively AI hub in Stockholm.

Thanks for coming back to YC.