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
Partner with foundation model training teams to determine optimal data: mixtures, training curricula, objectives, and other design choices that shape model capabilities.
Design global-scale pipelines for dynamically updating and reusing data: mixtures as our understanding of model behavior evolves and real-world datasets are continually ingested.
Develop and deploy principled data-selection algorithms (e.g. influence: estimation, proxy models, gradient similarity search) to find targeted subsets of data for fine tuning and RL.
Run large-scale empirical studies linking pre-training data composition to: downstream post-training performance, translating those scaling laws into optimized production training runs.
Collaborate closely with behavior, perception, and evaluation teams to test: these foundation models and measure their impact on the performance of the Waymo Driver.
PhD in ML or related in AI/CS, or MSc with a strong academic publication record
What they're looking for
- Strong statistical / ML knowledge
- Proficiency in Python
- + years of experience with modern deep learning frameworks: JAX, TensorFlow/PyTorch
- Proven track record of deploying ML models to production environments