Senior RL Scientist: Autonomous Planning & World Modeling

Waymo; Kirkland (WA); Hybrid; USD 204,000 - 259,000; Full time

Benefits offered by this job

Health, dental, vision insurance

401(k) with company match

Paid time off

Job summary

Waymo AI Foundations is seeking a highly skilled researcher to advance reinforcement learning and foundation models for autonomous vehicle trajectory planning. In a hybrid role, you will report to a Principal Scientist and work across teams to bring cutting-edge methods from research into production pipelines.

You will develop scalable training flows, contribute to post-training evaluations, and publish impactful results while collaborating with distributed systems teams to optimize inference at

Qualifications

• PhD or Masters in Computer Science, Machine Learning, Robotics; with 3+ years of industry or post‑doc research in RL or Foundation Models.
• Demonstrated contributions to the field via publications, blogs, or open-source projects.

Responsibilities

• Participate in Waymo’s Foundation World Model post‑training and evaluation.
• Research and develop RL and Distillation techniques for Autonomous Vehicle Trajectory Planning.
• Integrate research into Waymo’s RL infra and run rigorous ablations to scale promising methods.
• Collaborate with engineering and research teams to share recipes and best practices.

Skills

Reinforcement Learning

Foundation Models

Multi-Modal learning

Distributed training

Model training flows

Data parallel

Tensor parallel

Education

PhD in CS/ML/Robotics

Masters in CS/ML/Robotics

Tools

Data parallel

FSDP

Tensor parallel

Job description

Waymo AI Foundations is seeking a highly skilled researcher to advance reinforcement learning and foundation models for autonomous vehicle trajectory planning. In a hybrid role, you will report to a Principal Scientist and work across teams to bring cutting-edge methods from research into production pipelines.

You will develop scalable training flows, contribute to post-training evaluations, and publish impactful results while collaborating with distributed systems teams to optimize inference at