Software Engineer - Dexterous Manipulation
Core
Implement, tune, and deploy reinforcement learning and learning-based control for high-DOF, multi-fingered robotic hands to enable human-like precision manipulation.
Role type
Senior IC software engineer (dexterous manipulation)
Builds
Production-grade manipulation software for humanoid robot hands running in simulation and on physical hardware
Domain
Robotics, Applied AI, Humanoid robots
Deliverable
production ML models
Required skills
Reinforcement learning for robotic control, Imitation learning, Diffusion policies, Python, C++, Physics simulation (IsaacSim, MuJoCo, Drake), Robotics kinematics and dynamics, Jacobian-based control, Sim-to-real pipeline development
Preferred skills
Teleoperation (VR/haptic), Motion retargeting, Tactile-sensing integration, Computer vision (6D pose, point clouds), End-effector calibration
Technologies
IsaacSim, MuJoCo, Drake, C++, Python
Responsibilities
Implement and tune control algorithms for multi-fingered hands (grasping, in-hand manipulation, tactile feedback), Develop low-latency manipulation software within real-time controls stack, Translate research methods into production-grade software, Build and refine sim-to-real pipelines, Deploy and debug manipulation capabilities on physical robots, Collaborate with hardware and systems engineers on sensor/actuator requirements
Seniority
Senior, hands-on IC
