Lead Robotics & Hardware Engineer

The role

We build large-scale datasets to train humanoid robots and embodied AI — via teleoperation setups (including warehouse environments) and egocentric human-motion capture. We need a senior engineers who can own everything on the hardware and robotics side
: the platforms, sensors, and capture rigs that produce the data, and the decisions about what to build and buy. Software, data pipelines, and MLOps are handled by dedicated engineers — your job is to hand them clean, well-captured data from hardware you designed and stand behind.

• Hardware decisions, end to end. Select, spec, and sign off on all robotics and capture hardware — humanoid and manipulator platforms, whole-body controllers, and the full egocentric stack (head-mounted cameras, VR headsets, motion capture, EMG and other wearables). You make the build-vs-buy calls.
• Teleoperation systems. Design and stand up teleop stations for warehouse and lab collection — rigging, sensing, real-time control, and the operator experience.
• Egocentric capture rigs. Own the design of wearable/head-mounted capture setups: sensor placement, field-of-view and coverage, calibration, and multi-modal time synchronization at the hardware level.
• Data quality at the source. Define capture standards (sync tolerances, FOV/coverage, command smoothness) and make sure the hardware meets them before data reaches the software team.
• Technical leadership. Set the hardware/robotics direction, review work, and be the senior voice the team relies on — including adjacent areas like RF/SDR and signal processing.

The strongest signal is demonstrated depth, not a year count — someone who has personally built and operated real robot and capture hardware in the field and owned the messy parts (calibration drift, sync offsets, hardware failing mid-collection).

• 8+ years in robotics/hardware engineering (or equivalent depth), hands-on with physical systems — not simulation alone.
• Proven end-to-end ownership of robotics or data-capture hardware that ran at real scale.
• Strong in robot kinematics, control, and teleoperation
  , and comfortable specifying sensors, cameras, and capture rigs.
• Hands-on with multi-modal sensor synchronization and calibration
  .
• ROS / ROS2 and the surrounding robotics tooling.
• Evidence of technical leadership — setting direction and raising standards.

Humanoid platforms (e.g. Unitree) and whole-body control egocentric / VR capture (PICO, Quest) and wearables (EMG) NVIDIA Isaac Sim / Isaac Lab / GR00T imitation learning / VLA models DSP, signal processing, or RF/SDR.