AI / ML Researcher

You push on the hard problems sitting between what models can do today and what customer operations actually need. More science than service: you set the agenda, run the experiments, publish where it makes sense, and hand frontier techniques to the engineers who ship them in six weeks.

What you own. A research agenda across the problems we keep hitting at the edge - agent reliability, long-horizon planning, retrieval over messy operational data, domain-specific evals. Rigorous experiments. Prototypes that cross the gap into production. A tight loop with the Technology Partners and Platform team - what we learn in a construction office is your best dataset.

What you bring. Deep ML background; PhD or equivalent industry research. Track record on applied problems, not just benchmarks - you've shipped research into something a real user touches. Strong engineering fundamentals; you don't throw work over a wall.