29 days ago
Base Salary
$251k - $310k/yr
Responsibilities
- Lead the technical strategy for curating and constructing massive-scale multimodal pre-training datasets and domain-adaptation data mixtures.
- Design and implement SFT and reinforcement-learning pipelines using RLHF, RLAIF, DPO, GRPO, and PPO to improve instruction following and reasoning.
- Architect scalable inference and evaluation pipelines for a VLM data flywheel that sources and autolabels driving edge cases.
- Conduct ablation studies and define training recipes, scaling laws, architectures, token budgets, and loss functions for multimodal training.
- Drive cross-functional AI strategy across ML Infra, Perception, Behavior, and AI Foundation teams.
- Own the foundation-model technical roadmap, lead design reviews, establish engineering practices, and mentor senior engineers.
Requirements
- Master’s degree in Computer Science, AI, ML, or a related technical field.
- 8+ years of hands-on experience designing, training, and scaling deep learning models, including at least 3+ years focused on LLM or VLM training.
- Expertise across the foundation-model lifecycle, including pre-training data curation, interleaved formats, tokenization, distributed training, and post-training.
- Expert-level understanding of FSDP, Megatron, JAX/Pax, or comparable distributed training paradigms.
- Expert software engineering fundamentals in Python, PyTorch, or JAX, with experience building reliable and scalable ML systems.
- Ability to define technical roadmaps and lead complex, multi-quarter initiatives across teams.
- Preferred qualifications include a PhD, strong publications in venues such as NeurIPS, ICML, ICLR, or CVPR, advanced reinforcement-learning experience, foundation-model data engineering at billion- or trillion-token scale, and robotics or autonomous-driving experience.
- Demonstrated Staff-level impact through product influence, zero-to-one ML architecture development, and technical leadership.
Benefits
- Hybrid full-time position available across US locations.
- Eligibility for a discretionary annual bonus program, equity incentive plan, and company benefits program, subject to eligibility requirements.
Categories
AI ResearchML Engineering
About Waymo
Waymo develops the Waymo Driver, a full-stack autonomous driving system, and operates the Waymo One robotaxi service that offers paid, driverless rides in select U.S. cities. The company monetizes through ride-hailing fares and partnerships integrating its technology on multiple vehicle platforms. Founded in 2009 as Google’s self-driving project, Waymo is headquartered in Mountain View and is an Alphabet subsidiary, with over 10 million rider-only trips and 100+ million autonomous miles on public roads.