• Design, implement and own ML metrics and evaluation pipelines spanning offline model evaluation, simulation and on-road performance. • Build and maintain test, regression and hillclimbing suites that gate model and stack releases, including automated triage of regressions to root cause. • Drive model improvement through loss analysis, error mining, and data balancing/curation strategies for training and evaluation sets.

• Deploy and Optimize Machine Learning model architectures across May’s Autonomous Driving training and inference stacks. • Own the model-compilation and deployment pipeline end-to-end. • Establish and defend latency/throughput budgets across the AV stack, including profiling, regression and integrity tests.

Role Description The Perception team builds the system which learns the spatial-temporal representation and their semantic meanings of the surrounding environment of the autonomously driving vehicle (ADV). We work jointly with downstream teams on the optimization and integration into the Waymo Driver. We conduct our own research to address real-world problems and collaborate with research teams at Alphabet. - Design, implement, and optimize large-scale continual pre-training pipelines for cutting-edge VLM foundation models. - Conduct research and development on novel pre-training techniques, focusing on efficiently integrating new, diverse, and multimodal data streams (e.g., visual data from different sensors) into existing models. - Develop and rigorously evaluate metrics and methodologies for measuring the performance, and transferability of continually pre-trained foundation models in the context of autonomous driving. - Stay current with the latest advancements in large language models, vision-language models, and continual learning, and translate relevant research into production-ready systems. Qualifications - 5+ years of experience in Machine Learning, with a focus on large-scale model development (LLM, VLM, or similar foundation models). - Proven expertise in LLM/VLM pre-training, continual learning with large scale datasets. - Strong coding proficiency in Python and deep learning frameworks (e.g., Jax, TensorFlow, PyTorch). - Hands-on experience with model training, evaluation, and deployment in a production environment. - Master's degree in Computer Science, Electrical Engineering, or a related field, or equivalent practical experience. Requirements - Experience in fine-tuning foundation models for autonomous driving or robotics applications. - Familiarity with large-scale data curation and quality assurance processes for multimodal datasets. - Background in autonomous vehicle perception, motion planning, or decision-making systems. - Publications in top-tier machine learning or computer vision conferences (e.g., NeurIPS, ICML, CVPR, ICCV, ECCV). - PhD in a relevant field. Benefits - Waymo employees are eligible to participate in Waymo’s discretionary annual bonus program. - Equity incentive plan. - Generous Company benefits program, subject to eligibility requirements. Salary Range The expected base salary range for this full-time position across US locations is $170,000 — $216,000 USD. Actual starting pay will be based on job-related factors, including exact work location, experience, relevant training and education, and skill level.

• Work as a product owner to deliver on the product vision and feature priorities. You should have a strong track record of making tough tradeoffs to balance scope, quality, supportability, performance, and time criticality. • Guide the team through design/implementation for complex technical projects. • Work closely with the internal stakeholders to ensure the product meets quality/stability requirements of enterprise customers, leveraging experience inventing and improving technology of performance/stability/scale. • Manage end to end product ownership, from planning and design to on call product support. Manage/fix/communicate issues that arise during escalations/customer issues.