SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Zoox, maker of the world's first purpose-built commercial robotaxi, is seeking a Software Engineer specializing in GPU compute performance. The Planner Compute team is responsible for optimizing the motion planner—the largest single software component in the autonomy stack—where low latency and consistent resource utilization are critical to safe operation.
You will be the team's expert in GPU performance, responsible for instrumenting, monitoring, analyzing, and optimizing GPU-based algorithms across the full scope of Zoox's compute needs: from machine learning inference to custom CUDA kernels written specifically for the motion planner.
Key responsibilities:
- Analyze performance metrics to identify GPU hotspots and optimization opportunities
- Adapt the current planner to multiple GPU architectures with different resource constraints
- Optimize ML models for latency and GPU memory usage
- Serve as a subject matter expert on CUDA and GPU performance, supporting other engineers within the Planner team
This is a senior individual contributor role where you will guide technical direction and mentor peers, but do not manage a team.
Requirements:
- BS in Computer Science or related field with 6+ years of professional experience
- Strong knowledge of CUDA as applied to recent GPU microarchitectures (e.g., Ampere, Blackwell) and hands-on experience debugging and optimizing GPU kernels using tools like Nsight
- Strong C++ proficiency and experience working in large codebases; comfortable in Linux development environments
- Experience developing, debugging, and profiling complex multiprocess systems (e.g., robotic systems, game engines)
- Bonus: GPU kernel development experience
