We are a premier value-added distributor and software developer for Unitree Robotics, working with the world's most advanced quadruped and humanoid robots. We are looking for a Robotics Platform Engineer to build the backbone of our new Teleoperation and Training Platform. You will not be designing robot control theory or building UI frontends. Instead, you will build the critical infrastructure that connects the robot to the cloud. You will be the "bridge" between our United States based robotics engineers (who handle the hardware control) and our India based web developers (who build the user interface). Your code will live on the robot, in the cloud, and in the network layer between them.
Key Responsibilities:
?
Containerize the Robot Stack:
? Wrap Unitree native libraries (unitree_sdk2, unitree_ros2) and our custom Python agents into highly optimized Docker containers.
? Ensure these containers run reliably on edge hardware (NVIDIA Jetson/Orin) and handle resource allocation (GPU passthrough) correctly.
?
Develop the "Web-to-Real" Bridge:
? Architect the low-latency signaling layer using WebRTC and WebSockets to allow browser-based control of the robot.
? Configure and tune DDS (Data Distribution Service) middleware to ensure reliable message passing between robot processes and the external network.
?
Build the Data Pipeline (The Moat):
? Create an automated pipeline that streams high-frequency telemetry and video from the robot to our cloud storage.
? Implement data post-processing scripts to automatically format logs into Hugging Face LeRobot standards for Imitation Learning.
?
Simulation-First Development:
? Maintain a CI/CD pipeline where code is automatically tested against unitree_mujoco or unitree_sim_isaaclab environments before deployment to physical hardware.
The Tech Stack
? Languages: Python (Primary), C++ (for SDK/middleware wrapping).