Let’s talk about a San Francisco outfit that decided the way personal injury law firms work is, well, broken. Not outdated. Not inefficient. Broken. Founded in 2022 by Pedro Paulino and Vitor Vavolizza, Tavrn didn’t come from casebooks or boardrooms; it came from deep inside the grind. Pedro, a Harvard dropout and competitive chess player, studied the board, saw the patterns, and knew exactly which pieces weren’t pulling their weight. Vitor, a self-taught coder since he was barely tall enough to reach the keyboard, brought the code and systems intuition most dev teams need a decade to fake.
Together, they built something bold: an AI-native platform that makes paralegal busywork beg for mercy. We’re talking medical chronology generation, demand letter drafting, record retrieval, client intake, done autonomously. Not assisted. Not augmented. Autonomous. Tavrn’s agents don’t just suggest; they deliver.
And the market noticed. Tavrn just closed a $15 million Series A, led by the heavyweights at Left Lane Capital, with support from A*, Hummingbird Ventures, and BoxGroup. That brings total funding to $21.6 million, not bad for a company that’s been in the game fewer rounds than a sparring partner.
But here’s what matters more than the money: momentum. With a client base of over 120 personal injury firms and an 80%+ drop in manual review time, Tavrn isn’t chasing efficiency, it’s cornering it. This isn’t about cutting hours. It’s about giving law firms the power to scale like tech companies, lean, fast, and lethal.
Now with Jon Parisi stepping in to lead sales, and engineering headcount about to balloon, Tavrn’s moving from the Bay’s backstreets to the national spotlight. Expansion plans include enterprise partnerships with top U.S. firms, a rollout of AI-powered eDiscovery, and eventual movement into adjacent verticals like medical malpractice. The ambition isn’t hidden, it’s built into the roadmap.
If you’re in legal tech or even orbiting it, this is the one to watch. Tavrn isn’t playing to catch up. They’re building the rails while driving the train. With models trained on real paralegal workflows, validated against actual production, and stacked on a cloud native backend that scales like a beast, this is what modern law looks like when AI actually knows what it’s doing.
