Member of Technical Staff, Model Evaluation

Threeum

Threeum is a frontier company building physical AI. We design the model and the machine as one system, so the full loop of training, deployment, and improvement runs faster, more capable, and more autonomous. Our goal is to make frontier AI physical, and to make it widely accessible.

The Role

We are looking for an engineer-researcher to build the evaluation stack for models whose outputs are actions, and whose failures happen in rooms rather than benchmarks.

Some example areas you might work on (not limited to):

• Design evals that predict deployed behavior, not just leaderboard behavior
• Build regression harnesses that catch capability and safety drift before release
• Define the evidence bar a model must clear to touch real hardware

If you want to build the machines the models deserve, we would love to hear from you.

We offer a base salary of $250,000 to $400,000 USD and a meaningful equity grant, depending on experience and background, along with competitive benefits.