AI runtime + evals engineer
Full timeRemote, LatamReports to Javier, our CTO
We're building the infrastructure that connects the world's capital with the world's risk. Our ambition is to make that connection a catalyst for financial innovation at scale.
For insurance entrepreneurs, that means a chance to turn ideas into programs and build new companies. For capital providers, it means better ways to deploy capital and participate in the economy of the future. We want to help more businesses come to life, and capital reach the risks it should be backing.
We're building two products to make that vision real. An MGA operating system to help MGAs launch and scale extremely efficient insurance programs. A risk transfer marketplace where MGAs and capacity providers meet, with a transparency layer that lets risk and capital flow freely through the economy.
You'll report directly to our CTO. This is a role with impact and growth. You'll have real ownership and real accountability. You'll be part of one of the best teams in the world, applying agentic AI in one of the largest industries. Our agents run in production for large clients. Your job is to make them reliable. You own the runtime they run on and the evaluation system that tells us, with evidence, whether they're getting better or worse. When a model provider fails at 3am, when a run breaks in a way nobody has seen before, when a change looks good but the evals say otherwise, you're the person who owns that.
We're hackers. We find ways to solve hard problems. We work with intensity, respect, honesty, kindness. That matters when you're working this hard. But this is not a job for everyone. It's demanding, uncertain, and intense. If you want to live the race of a high growth startup, aiming to build something big, this might be the best decision you ever make.
Email javier@gobraven.cc with the role name in the subject. Tell us about an agent you've run in production and how you knew it was working.