Tech Lead (Robotics Runtime)

Location: Palo Alto, United States; on-site

Company

Building generalized physical AI and robotic systems for dexterous, adaptive, reasoning-intensive work in real-world industrial environments.

What you will do

• Design and build the robotics runtime and middleware layer, including message architecture, QoS policies, and multi-process communication.
• Own onboard inference serving for VLA and action-expert models, including batching, quantization, hardware acceleration, and latency tradeoffs.
• Architect multi-sensor synchronization and fusion across different clocks and cadences for closed-loop control.
• Make technical decisions on middleware, process and thread architecture, real-time guarantees, and compute hardware.
• Set technical standards, conduct deep design and code reviews, and prototype new runtime, hardware, and model-serving architectures.
• Work directly with research and modeling partners to translate policy requirements into runtime architecture.

Requirements

• Strong systems programming skills in Python and/or Rust.
• Hands-on robotics middleware experience with ROS2, DDS implementations such as Fast DDS or CycloneDDS, or Zenoh; including custom messages, QoS tuning, and multi-node communication.
• Experience with real-time systems, scheduling, latency budgets, jitter, RTOS concepts, and embedded or edge Linux.
• Experience with sensor time synchronization, PTP or gPTP, sensor fusion, and clock drift or skew across distributed robot compute.
• Experience deploying and optimizing models with ONNX Runtime, TensorRT, or similar tools, including quantization and batching for edge inference.
• Experience with control systems, motion primitives, and taking real-time software from prototype to reliable operation on physical hardware.

Nice to have

• Familiarity with Isaac Sim, MuJoCo, or Gazebo.
• Exposure to functional safety concepts for physical systems.
• Background in autonomous vehicles, industrial IoT, or similar real-time and embedded domains.
• Experience scaling runtime platforms across many concurrent production robots.
• Prior technical mentorship or informal lead experience without formal management responsibility.

Culture & Benefits

• Hands-on work with physical systems and real-world industrial robotics.
• IC-heavy ownership of difficult real-time robotics and model-serving problems.
• Direct collaboration with research and modeling partners.
• Focus on balancing rapid experimentation with runtime reliability, determinism, fault tolerance, and safety margins.