7+ years of experience, including 4+ years in large-scale model training platforms.
Deep knowledge of distributed training frameworks like Megatron-LM and TorchTitan.
Experience with reinforcement learning frameworks and algorithms such as GRPO and PPO.
Familiarity with collective communication libraries like NCCL for optimizing multi-node training.
Strong skills in profiling tools like NVIDIA Nsight and PyTorch Profiler to identify bottlenecks.
Proficient in Python or C++.
Experience with ML compilers and hardware/software co-design.
Responsibilities
Lead design and implementation of training performance optimizations for VLM and VLA models.
Own technical strategy and roadmap for reinforcement learning performance optimization.
Drive profiling and root-cause analysis of throughput bottlenecks across the stack.
Collaborate with ML researchers and engineers to define requirements for training infrastructure.
Build custom kernels and benchmarks to validate performance gains.
Mentor engineers and elevate the team's technical standards through design reviews.
Benefits
Opportunity to work on cutting-edge autonomous driving technology.
Collaborative environment with cross-functional teams.
Mentorship opportunities to shape the next generation of engineers.
Access to advanced ML infrastructure and tools.
Potential for career growth in a rapidly evolving field.
Full Job Description
The OpportunityAre you excited to own ML training performance at Zoox and make the models that power autonomous driving train faster, cheaper, and at greater scale? You'll set technical direction for distributed training, parallelism strategies, mixed-precision training, and RL on SOTA accelerators, especially for multi-modal VLA and VLM models. You'll partner with all Autonomy teams (Perception, Prediction, Planner, Simulation, Collision Avoidance) and act as a force multiplier for our internal customers, while mentoring engineers and shaping the team's roadmap as we expand into new ML domains. If you want to learn more about our stack behind autonomous driving, please look here. If you want to learn more about our ML Infrastructure, here is one of our past talks at re:Invent.
In this role, you will:
Lead the design, implementation, and operation of training performance optimizations (data/tensor/pipeline parallelism, mixed precision, gradient checkpointing, GPU kernels) to scale VLM, VLA, and Foundation model training.
Own the technical strategy and multi-quarter roadmap for RL framework and RL performance optimization for driving foundation models.
Drive profiling and root-cause analysis of throughput bottlenecks across the compute, compiler, framework, and cluster-networking stack.
Partner with ML researchers, data engineers, and hardware engineers to define requirements and make architectural decisions for large-scale training infrastructure.
Build custom kernels and benchmarks that track and validate performance gains across the stack.
Mentor engineers and raise the team's technical bar through design reviews and best practices.
Qualifications
Note: You do not have to meet all the requirements below to be considered for this position:
7+ years of total experience, including 4+ years on large-scale model training platforms, with a record of leading projects end to end.
Deep experience with distributed training frameworks (Megatron-LM, TorchTitan) and efficiently utilizing GPU/accelerator clusters at scale.
Experience with RL frameworks (Verl/MILES) and algorithms (GRPO, PPO).
Experience with collective communication libraries (e.g., NCCL) and optimizing communication/compute overlap in multi-node training.
Strong command of profiling tools (NVIDIA Nsight, PyTorch Profiler, or XLA/JAX profiling) to find and resolve training bottlenecks.
Proficiency in Python or C++.
Experience with ML compilers, performance tooling, or hardware/software co-design (GPU/TPU/accelerator architectures).
Ability to influence across teams and communicate technical tradeoffs to both researchers and engineers.
$254,000 - $350,000 a year
About Zoox.com
Zoox is a self-driving car company that is developing a fully autonomous vehicle for ride-hailing services. The company was founded in 2014 and is headquartered in Foster City, California. Zoox's vehicle is designed to be electric, fully autonomous, and capable of carrying passengers in a variety of seating configurations. The company is focused on developing a complete end-to-end solution for autonomous ride-hailing, including the vehicle, the software, and the infrastructure. Zoox has raised over $800 million in funding to date, and is backed by investors including Blackbird Ventures, Lux Capital, and DFJ.
