Zoox
Software Engineer, ML Training Optimization
$177K–$270K
Foster City, CA · On-site · Full-time · Posted Oct 16, 2026
- Experience
- 4+ years total, including 2+ years working on large-scale model training platforms.
About the job
Optimizes and operates large-scale ML training systems for Zoox’s autonomous-driving foundation, VLM, and VLA models.
What you'll do
- Implement training-performance techniques, custom kernels, and benchmarks to improve model-training efficiency.
- Optimize reinforcement-learning frameworks and profile bottlenecks across compute, compiler, framework, and cluster networking.
- Collaborate with research, data, and hardware teams on requirements and architecture for large-scale training infrastructure.
What they ask for
- 4+ years of total experience, including 2+ years working on large-scale model training platforms.
- Experience with distributed training frameworks and using GPU/accelerator clusters at scale.
- Experience with RL frameworks and algorithms, collective communication libraries, and profiling tools to resolve training bottlenecks.
- Proficient in Python or C++.
Nice to have
- Experience with ML compilers, performance tooling, or hardware/software co-design.
- Experience with Megatron-LM, TorchTitan, Verl/MILES, GRPO, PPO, NCCL, NVIDIA Nsight, PyTorch Profiler, or XLA/JAX profiling.
Pay
USD 177,000–270,000 per-year-salary (base salary)
Tools
Megatron-LM, TorchTitan, Verl/MILES, GRPO, PPO, NCCL, NVIDIA Nsight, PyTorch Profiler
AI focus
ML training optimization and reinforcement-learning performance for autonomous-driving foundation, VLM, and VLA models.
Model-written summary · the original posting is authoritative