ML Engineer, Manipulation
Core
Develop and deploy learning-based manipulation systems enabling mobile robots to interact reliably with the physical world in dynamic human environments.
Role type
Senior IC machine-learning engineer (robotic manipulation)
Builds
Perception-to-action models, training datasets, evaluation tooling, and deployment pipelines for service robots
Domain
Robotics, AI, Physical Interaction
Deliverable
production ML models
Required skills
PyTorch, Python, training/evaluation pipelines, robotics manipulation, visuomotor control, sequential-to-sequence models
Preferred skills
Vision-Language-Action (VLA) models, behavior cloning, transformer/diffusion policies, sim-to-real training (Isaac Sim/Mujoco), edge deployment (ONNX/TensorRT), safety-critical robotics integration
Technologies
PyTorch, Python, Isaac Sim, Mujoco, ONNX, TensorRT
Responsibilities
Develop learning-based manipulation models for end-to-end sensor-driven interaction; Build and maintain manipulation training pipelines including dataset creation and distributed training; Design evaluation metrics and regression tests for manipulation reliability; Develop sim-to-real workflows including simulation environments and domain randomization; Optimize and distill models for edge deployment; Partner with AI platform team to integrate policies with control and safety systems; Analyze field performance and drive iterative improvements through data collection
Seniority
Mid-to-Senior, hands-on IC
