We work on the dexterity problem for humanoids

“Join our platform team to build the data infrastructure that powers our robotic learning systems. You'll develop high-performance logging pipelines, build visualization tools for multimodal sensor data, and create systems that help our AI teams train better robotic policies.”

🎯Role Details

• Design and implement high-throughput data logging systems for real-world robotic platforms
• Build visualization and analysis tools for multimodal sensor streams (vision, tactile, force, proprioception)
• Develop sensor calibration, time-synchronization, and validation pipelines
• Create data management and tooling to support large-scale AI and robotics training workflows
• Work with cloud infrastructure to enable scalable data storage, processing, and retrieval
• Collaborate closely with robotics, perception, and learning teams to improve data quality and usability
• Support both real-robot and simulation-based data collection pipelines

⚡Requirements

• BS or MS (or equivalent experience) in Computer Science, Robotics, or a related engineering field
• Strong experience with Linux systems, shell scripting, and low-level system tools
• Familiarity with performance profiling and debugging (e.g., perf, strace, gdb)
• Experience building data pipelines, logging systems, or visualization tools
• Understanding of sensor calibration, synchronization, and data storage formats
• Proficiency in Rust (C/C++, Python are nice addition)
• Familiarity with ROS / ROS2, bag files, and common robotics data formats
• Experience with robotics simulators such as MuJoCo or Isaac Sim
• Familiarity with URDF / MJCF / SDF and simulation-robot interfaces
• Exposure to cloud platforms (AWS, Azure) or SIL/HIL testing setups

Nice to Have

• Experience with high-frequency or real-time data systems
• Background in distributed systems or large-scale data processing
• Interest in robotics learning, perception, or manipulation

Robotics hardware + data collection + AI