About the job
Waymo LLC develops autonomous driving technology in Mountain View, CA, with a mission to improve road safety and mobility. Since its origins as the Google Self-Driving Car Project in 2009, Waymo has logged over 100 million miles on public roads and completed more than ten million rider-only trips across more than 15 states.
Perception Team
The Perception team builds the systems that allow the Waymo Driver to understand its surroundings, make decisions in real time, and ensure passenger safety. This work relies on large-scale, multi-modal machine learning models trained on millions of miles of sensor data.
Role overview
This Principal Software Engineer (hybrid) will report directly to the Senior Director of Engineering, Perception. The position focuses on advancing the Waymo Driver’s post-training capabilities and supporting the technology roadmap for perception models.
Key responsibilities
- Partner with Onboard model engineers to prototype and test new model designs and training methods before they move to production.
- Create infrastructure for model experimentation, including robust data and evaluation frameworks.
- Lead the post-training technology roadmap, exploring reinforcement learning and related techniques.
- Support production rollout of major improvements in model architecture and training recipes to address diverse behavioral situations.
- Foster strong partnerships between Onboard and research teams within Waymo and Alphabet.
- Collaborate with Onboard and other stakeholders to evolve and execute the post-training recipe roadmap for significant advancements and production impact.

