Staff Machine Learning Engineer – VLM/LLM Evaluation

The mission of the Waymo AI Foundations team is to develop machine learning solutions addressing open problems in autonomous driving, towards the goal of safely operating Waymo vehicles in dozens of cities and under all driving conditions. As part of our work, we also initiate and foster collaborations with other research teams in Alphabet. AI Foundations areas that we are currently focusing on include reinforcement learning, learning from demonstration, generative modeling, Bayesian inference, hierarchical learning, and robust evaluation.

This role follows a hybrid work schedule and you will report to a Senior Staff Software Engineer.

You will:

• Work with a creative team of people who help to build the state-of-the-art Foundation Models that are used throughout Waymo’s systems, both onboard autonomous vehicles and offboard in simulation
• Lead the development of end-to-end evaluation systems and benchmarks for Waymo Foundation models, encompassing the entire lifecycle from pretraining and supervised fine-tuning (SFT) to reinforcement learning (RL), for evaluating the quality, safety, and realism of embodied AI agents
• Partner within and across organizations to land disruptive and innovative tech in production
• Implement and extend large large scale data and evaluation pipelines

You have:

• Master’s degree or PhD degree in Computer Science, similar technical field of study, or equivalent practical experience
• 5+ years of experience in ML engineering and applied Deep Learning, with a strong portfolio of shipped products or publication record
• Experience with large scale distributed system
• Proficient programming skills (eg: Python, C/C++)
• Strong analytical and debugging skills

We prefer:

• ML infra experience: training, evaluating and deploying ML models at scale
• Deep learning experience, especially with generative models, e.g., LLMs/VLMs, and/or reinforcement learning 
• Proficiency and in-depth knowledge of the inner workings of an ML framework (e.g. Pytorch, JAX, Tensorflow)