Machine Learning Engineer, Robot Learning, Loco-Manipulation

What this role involves

Path Robotics is building a new Robot Learning team focused on whole-body loco-manipulation. The role involves building the robot-learning stack from the ground up, including training infrastructure, simulation, and deployment. The candidate needs hands-on robot learning experience with sim-to-real transfer, diffusion/flow-matching policies, and reinforcement learning on real hardware.

Skills this posting asks for

• robot learning
• sim-to-real transfer
• diffusion-based action policies
• flow-matching action policies
• action chunking
• reinforcement learning for robotics
• python
• pytorch
• ml training infrastructure
• nvidia isaac sim
• nvidia isaac lab
• mujoco
• physical robot debugging
• edge inference
• tensorrt
• onnx
• fp16 quantization
• int8 quantization
• visual self-supervised representation learning
• behavioral cloning
• data pipelines
• experiment tracking
• data versioning

Requirements

• 2 years of experience

From the employer’s posting

Build the Path Forward At Path Robotics, we’re building the future of embodied intelligence. Our AI-driven systems enable robots to adapt, learn, and perform in the real world closing the skilled labor gap and transforming industries. We go beyo…