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

Legora brings document review, drafting, and research into a workspace with a web app and a Microsoft Word add-in. Strand says GPT-3.5 made it practical to build tools for unstructured legal documents, and the company grew from an MVP into a system used by tens of thousands of lawyers daily.

Strand describes due diligence grids that check many documents against questions and link answers to citations. Legora also offers agents for multi-step tasks, such as researching and drafting a memo, and playbooks that review contracts against rules and suggest fallback language. He says scaling these workflows requires careful document search, chunking, and reliable citations.

Legora positioned itself as a long-term partner whose success is tied to law firms' success. Strand says he interviewed 100 lawyers, learned from daily feedback, and worked with early partners to understand different legal practices. For larger firms, he recommends starting with senior people or one partner's team, then expanding as others see the results.

Strand says generative AI made it faster to ship products and challenged older legal tech point solutions. Legora grew from about 10 people after Y Combinator to around 100, while expanding across Europe and into the US. Its system can switch among models from OpenAI, Anthropic, Google, and Mistral, with simpler models handling simpler queries to manage cost.

Strand expects lawyers to spend more time directing and checking AI agents while remaining responsible for work delivered to clients. He advises vertical AI founders to avoid dependence on one model provider, understand where they add lasting value, and use models creatively for specialized work. Legora's own direction, he says, is to combine software with strategic support as firms adapt.