The opportunity
Waymo is an autonomous driving technology company with the mission to be the world's most trusted driver. Since its start as the Google Self-Driving Car Project in 2009, Waymo has focused on building the Waymo Driver—The World's Most Experienced Driver™—to improve access to…
What you'll do
Design, implement, and benchmark a system where LLM meta-agents generate,: synthesize, and iteratively refine reward functions for autonomous driving planning policies
Diagnose policy failure modes, reward exploitation, and multi-objective: trade-offs, integrating findings back into the automated discovery loop
Collaborate with scientists engineers to transfer validated reward: formulations and tooling into planner recipes
Document research outcomes, distill key design patterns, and prepare findings: for internal tech transfers and potential conference publication
Currently pursuing a PhD in Computer Science, Robotics, Machine Learning, or a related quantitative field
Strong hands-on experience in Reinforcement Learning (RL / RLFT) and training: deep learning policies with PyTorch or JAX
What they're looking for
- Demonstrated proficiency in Python and modern ML engineering practices
- Familiarity with autonomous vehicle planning, trajectory optimization, or: sequential decision-making under uncertainty
- Track record of peer-reviewed publications at top AI/robotics conferences: (e.g., NeurIPS, ICRA, IROS, CoRL, ICML, ICLR, CVPR)
- Experience with LLM-guided code generation, automated reward discovery (e.g.,: Eureka-like frameworks), or meta-RL methods
