Embodied AI Systems Engineer
Mission
Take embodied AI from hypothesis to a system that runs in real care settings.
The first problem you'll tackle
In 6–12 months: ship a testable perceive–understand–act loop and prove at least one task end-to-end.
Core responsibilities
- Own the system architecture and validation plan for the first tasks.
- Wire together the robot platform, perception, models, tasks, and the eval loop.
- Hunt failure modes in sim and the real world, and iterate fast.
What we look forNo one masters every area — we care more that you go deep in one and are willing to cover the rest of the system
- Research or engineering background in embodied AI, robotics, or multimodal AI.
- You turn research into testable, iterable prototypes.
- Strong systems thinking, self-direction, and a knack for decomposing hard problems.
- Deep in at least one of robotics, multimodal AI, task planning, simulation, or control — and willing to cover the rest.
Nice to have
- VLA / VLM, AI agents, RL, simulation training, or sim-to-real.
- Real-robot debugging, published papers, open source, or shipped prototypes.