Staff Machine Learning Engineer, AI Infrastructure

Waymo•San Francisco, NY

About The Position

Waymo’s Driver Understanding and Evaluation (DUE) team develops metrics and technologies to understand the Waymo Driver’s real-world behavior, including context and scene analysis, real-world driving data augmentation, large-scale data infrastructure, and improvements to agents and simulation. The DUE Machine Learning team builds and operates scalable machine learning and data systems, simulation workflows, and insight tools to improve evaluation and developer workflows. It combines expert human judgments with advanced machine learning models to deliver training and evaluation data for the many metrics and components that make up the Waymo Driver. The Staff Machine Learning Engineer will lead development of machine learning infrastructure spanning data, training, and production serving. The role includes scaling distributed systems for dataset generation, training, and evaluation; ensuring production-serving reliability and performance; maintaining consistency between offline evaluation and production inference; setting technical direction; mentoring engineers; and collaborating across modeling, evaluation, and platform teams.

Requirements

Bachelor’s degree in Computer Science, Machine Learning, or a related technical field, or equivalent practical experience.
At least 7 years of hands-on experience building machine learning systems.
Experience designing, scaling, training, deploying, and optimizing large-scale ML systems from data to model.
Strong coding experience in Python and/or C++.
Experience with ML frameworks such as JAX, TensorFlow, or PyTorch, including user-facing APIs and internals.
Proven ability to set technical direction and influence partner teams in ambiguous problem spaces.

Nice To Haves

Master’s or Ph.D. in Computer Science or a related field, specializing in machine learning.
At least 10 years of hands-on experience building production machine learning systems.
Experience serving ML models in production at scale, including low-latency inference on TPUs or GPUs, SLOs, and safe model rollouts.
Technical leadership serving production-quality models, including owning launches, reliability, and on-call for services other teams depend on.
Experience with distributed training and accelerator performance profiling.
Experience in autonomous vehicles, robotics, or large-scale simulation.

Responsibilities

Lead the design and development of ML infrastructure taking DUE models from data through training to production serving, including event detection models, triage foundation models, and agents.
Design and scale distributed systems for large-scale dataset generation, model training, and evaluation across real-world and simulation data.
Own reliability, efficiency, and observability of production ML serving, including accelerator capacity planning and performance optimization.
Keep offline evaluation and production inference consistent.
Drive technical direction, guide architectural decisions, and mentor engineers.
Collaborate with modeling, evaluation, and platform teams to translate product and business goals into measurable technical deliverables.