Software Engineer

Ambitus designs the reinforcement-learning environments that the world's top AI labs train on — the curriculum behind the next generation of frontier models.

We work directly with the largest labs and ship the environments that move their benchmarks.

What you'll do

• Build new RL environments from scratch — design the task, write the simulator, ship the evals
• Work alongside lab researchers to understand what their models can't do yet, then build environments that close that gap
• Own end-to-end: protocol, infra, scoring, deployment

What we look for

Strong slope above all else. We hire on what you can build today versus a year ago, not on pedigree. Bring evidence — projects, ships, things you made that didn't exist before you.