Senior Research Scientist, World Action Modeling

Job Summary

The Waymo AI Foundations team is seeking a researcher to design and implement generative world action modeling solutions. This role involves developing scalable data pipelines, evaluating model behaviors, and collaborating across Waymo/Alphabet to apply these techniques to Waymo products.

About the Role

This role follows a hybrid work schedule and reports to a Staff Research Scientist / Tech Lead Manager. The mission of the Waymo AI Foundations team is to develop machine learning solutions addressing open problems in autonomous driving.

Responsibilities

• Design and implement generative world action modeling solutions to simulate future world states / observations and generate policies for embodied agents (autonomous vehicles).
• Develop and maintain scalable data pipelines to process data from multiple sources.
• Design and implement evaluation strategies for world action models.
• Study and analyze different behaviors of this model, such as scaling efficacy, downstream quality implications, model architecture design ablations, etc.
• Conducting cutting-edge research and potentially communicating research findings to the wider academic community via technical blog posts, technical reports and/or publications.
• Collaborate inclusively across Waymo/Alphabet to apply developed techniques to Waymo products.

Requirements

• PhD degree in Computer Science or a similar discipline, or an equivalent amount of deep learning research experience, with 5+ years of experience with Deep Learning and Generative Models.
• Strong coding skills in Python and strong familiarity with major ML Frameworks (JAX, Tensorflow, Pytorch).
• Strong familiarity with AI tools in day to day work.
• Experience in large-scale distributed training and different forms of parallelism.
• Experience in generative models for domains such as world models, images, videos, 3D, human animation, traffic, and/or simulation, using techniques such as diffusion and/or autoregressive models.
• Publishations at top-tier conferences like CVPR/ICCV/ECCV/ICLR/ICML/NeurIPS/IROS/CoRL etc.
• Substantial involvement in and contributions to high impact industry AI projects.
• First-hand experience in, and direct contributions to training large world models / world-action models.
• Familiarity with Reinforcement Learning in simulation environments.

Qualifications

• Strong theoretical background in machine learning and computer science.
• Experience with reinforcement learning in simulation environments.
• Experience with generative models for various domains.

Skills

• Strong coding skills in Python and strong familiarity with major ML Frameworks (JAX, Tensorflow, Pytorch).
• Experience with large-scale distributed training and different forms of parallelism.
• Experience with generative models for various domains.
• Familiarity with reinforcement learning in simulation environments.

Benefits

Benefits for this role include: Health, dental, vision, life, disability insurance, retirement benefits: 401(k) with company match, paid time off: 20 days of vacation per year, accruing at a rate of 6.15 hours per pay period for the first five years of employment, sick time: 40 hours/year (statutory, where applicable); 5 days/event (discretionary), maternity leave (short-term disability + baby bonding): 28-30 weeks, baby bonding leave: 18 weeks, holidays: 13 paid days per year.