Staff Research Scientist, Foundation Models Recipes
Core
Own data recipes for Waymo's Foundation Model pre-training and post-training, leading science on clustering, filtering, de-duplication, and long-tail data mining.
Role type
Staff Research Scientist (Foundation Models Recipes)
Builds
Data flywheel and data-centric evaluation techniques for Waymo's autonomous driving foundation models
Domain
Autonomous driving, Foundation Models, Data-centric AI
Deliverable
production ML models
Required skills
Data-centric AI, Reinforcement Learning, Foundation Models, ML framework internals (PyTorch/JAX/TensorFlow), rigorous ablation design, data mix optimization
Preferred skills
Large-scale foundation model data recipes, ML infrastructure at scale, Generative models (LLMs/VLMs), Deep learning
Technologies
PyTorch, JAX, TensorFlow
Responsibilities
Own data recipes for pre-training and post-training; Lead science on clustering, filtering, de-duplication, and long-tail data mining; Tech-lead research engineers to build the data flywheel; Integrate emerging research for scaling ladders and RL recipes; Engage with the research community on data-centric evaluation; Partner with engineering and research teams to share techniques
Seniority
Staff, hands-on IC with team leadership