Legal AI Annotator
Pubblicato il 17-10-2026 - US Health Partners in Venezia
At Lexroom, we set the bar for legal AI – clear, fast, and built for trust. In just a few years, what began as an idea has become a reliable platform adopted by thousands of legal professionals across Europe. Our mission is clear: make the law faster, clearer, and more reliable, giving lawyers, law firms, and companies the time and confidence to focus on what truly matters. We're scaling rapidly, backed by top-tier investors (having recently closed our Series B) and a world-class team dedicated to rigor, clarity, and impact. Today, Lexroom is a trusted partner, shaping the future of legal AI with discipline and purpose. Why Lexroom? Shape the Standard: Be part of the team setting the benchmark for legal AI in Europe. Work with the Best: Collaborate with exceptional colleagues in a culture that values competence, curiosity, and ambition. Deliver Meaningful Results: Your legal judgment will directly shape the quality of answers thousands of professionals rely on every day. At Lexroom, we leave buzzwords behind. Here, you'll build something that lasts: a trusted platform,
a strong reputation, a new standard for the legal world. The Role & Why Now The quality of our legal answers is the core of Lexroom's services. To measure and continuously improve it, we're looking for lawyers who can put their expertise to work evaluating AI systems. You’ll work independently and remotely on dedicated platforms, following Lexroom's guidelines, in direct contact with our People team. Key Responsibilities The scope of these activities may expand or change over time. Legal AI Benchmarks: You’ll evaluate AI-generated answers to legal questions. You’ll also define the positive and negative criteria used to judge whether an answer is correct. Finally, you’ll build and review the evaluation sets we use to measure the quality of Lexroom's services. Arena (Comparative Evaluation): You’ll compare alternative answers from different models to the same legal question. For each comparison, you’ll