Teams at Zoox

Lead the architecture and strategy for onboard GPU execution in Zoox autonomous vehicles. Define GPU scheduling, multi-tenancy, isolation, deterministic latency, compute budgets, hardware selection, and inference performance direction. Drive end-to-end reductions in latency, utilization, power, and cost across models, runtimes, middleware, and hardware. Provide technical leadership, mentor senior engineers, align cross-functional teams, and lead platform bring-up and performance optimization.

Lead ML training performance for large-scale VLM, VLA, foundation-model, and reinforcement-learning systems. Design distributed parallelism, mixed-precision strategies, gradient checkpointing, GPU kernels, profiling workflows, and performance benchmarks. Own RL optimization strategy and roadmap, diagnose bottlenecks across compute and networking layers, partner with research and engineering teams on architecture, and mentor engineers.

Design, train, and validate machine learning models that improve simulated LiDAR, radar, and camera realism for autonomous vehicle simulation. Build datasets from real and simulated sensor data, integrate 3DGS and NeRF reconstruction, develop generative augmentation methods, define sensor realism metrics, reduce sim-to-real gaps, and mentor engineers and interns.