Staff / Senior Staff Machine Learning Engineer - Collision Avoidance System

The Collision Avoidance System (CAS) is Zoox's independent safety layer. It runs alongside our main autonomy stack, constantly watches for imminent collisions, and reacts within milliseconds to prevent or reduce them. By using different methods and data paths than the main stack, CAS gives our robotaxi a second, diverse line of defense in the most challenging dense urban environments.

As a Staff or Senior Staff engineer, you will be the technical lead for CAS algorithms, with a focus on CAS perception. You will shape how CAS sees the world, combining interpretable geometric methods with machine learning, and you will own the architecture and roadmap that take these capabilities from idea to safe, real-time deployment on our fleet. Your work will directly affect the safety of every rider and road user around our vehicles.

In this role, you will: • Serve as the technical lead and architect for CAS perception, defining how lidar, radar, and camera data become reliable obstacle detection, tracking, and collision-risk estimates under tight latency and compute budgets

• Design, implement, and ship both machine learning and classical perception algorithms (e.g., point-cloud geometry, ground estimation, clustering, state estimation and tracking, learned detection and segmentation, sensor fusion) in production-quality, safety-critical C++

• Own the end-to-end CAS algorithm architecture and multi-release technical roadmap, making clear trade-offs between safety, performance, latency, compute, and ML vs. non-ML approaches

• Partner across CAS perception, behavior and planning, and validation to define interfaces, requirements, metrics, and evaluation strategies that show the system is safe and ready to launch

• Work with sensor, system performance, safety strategy, and main autonomy teams to drive cross-team technical decisions, while mentoring engineers and raising the technical bar across CAS

Qualifications • BS, MS, or PhD in Computer Science, Electrical Engineering, Robotics, or a related field, or equivalent experience

• At least 10 years of experience developing perception or related algorithms for robotics, autonomous vehicles, or other real-time systems

• Deep expertise in 3D perception from lidar, radar, and/or camera, with hands-on experience in both machine learning (detection, segmentation, fusion) and classical methods (geometry, state estimation, tracking)

• Strong C++ or Python skills and a track record of shipping efficient, reliable real-time software to production

• Experience leading the architecture and technical roadmap of a complex production system across multiple teams

Bonus Qualifications • Conference or Journal publications in Machine Learning or Robotics-related venues

• Experience with safety-critical systems or functional safety practices (e.g., ISO 26262, SOTIF), collision avoidance, emergency braking, prediction, or planning systems for autonomous vehicles

• Experience optimizing and deploying models or algorithms on onboard compute (e.g., GPU / CUDA, TensorRT)