← All open roles
ML Engineer, Simulation Infrastructure
Build the simulation environments and infrastructure that let the team iterate on hand-mocap models without waiting on real-world collection.
What you'll do
- Construct physically realistic hand and object interaction environments in simulation.
- Close the sim-to-real gap between policies trained in sim and the signal collected in the field.
- Scale up simulation data production for research and training.
- Collaborate with the AI and collection teams to develop new training and evaluation tasks.
- Prototype new sensor configurations in sim before they hit hardware.
What we're looking for
- 4+ years programming in Python and/or C++.
- Built environments in MuJoCo, PyBullet, Isaac Sim, or similar.
- Experience improving physics simulator or rendering performance.
- Discipline around testing - your sim stack stays correct as it grows.
- MS or PhD in CS or related field, or equivalent demonstrated impact.
Nice to have
- Calibration algorithms (extrinsic, intrinsic) for robotics or vision.
- Differentiable simulation experience.