Build the AI systems that have to be right.
We're a small team shipping AI into regulated work — insurance, healthcare, audit, public safety. The eval is the bar. The audit trail is the deliverable. The model is the easier half.
Hold all three views.
Modeling, product, and executive risk — in the same conversation. We hire engineers who can push back on a spec, an eval, and a board narrative without changing rooms.
Eval is the deliverable.
The model is the easier half. The eval that defends it for two years, the audit trail that replays a decision eighteen months later — that's the work.
Sustainable intensity.
Quarterly arcs with one on-site week, four weeks focused remote. No surprise weekends, no crunch as a permanent state. Regulated work demands intensity, not heroics.
3 roles open.
Founding ML Engineer
Set the modeling bar for everything we ship into regulated work.
Applied Research Engineer, Evaluation
Build the evaluation backbone that decides whether a model ships, stays shipped, or gets pulled.
Product Engineer
Design the surfaces operators actually trust — where the model's work shows itself.
Don't see the right role?
If you've shipped AI into regulated work and want to talk about something that isn't listed, write to us anyway. We open new positions when the right person turns up.