Senior Software Engineer, Model Evaluation - Simulation

The Simulator team builds state of the art simulations of realistic environments for the testing and training of the Waymo driver. We use machine learning to model the real world, including realistic agents (vehicles, pedestrians, cyclists, motorcyclists), roads, traffic control systems, and weather. To guide the development of simulator realism components, we create metrics to measure realism, which are used to assess the quality of the simulator, identify sim2real gaps, and used as loss functions to train a more realistic simulator - which in turn is used to assess the quality of the Waymo Driver.

In this hybrid role you will report to a Engineering Manager.

You will:

• Develop novel methods to measure sim realism in the context of several simulator tech and applications
• Develop efficient platform to allow measuring sim realism at scale
• Design systems that can automatically discover useful realism metrics to guide simulator development
• Collaborate with a world-class engineering and research team that use large models to improve sim realism
• Be a part of a collaborative engineering team that build a realistic simulator which powers the evaluation of the Waymo Driver

You have:

• Coding and design skills
• 5+ years of experience in evaluating ML models or complex systems, or building scalable infrastructure and pipelines for evaluation

We prefer:

• Navigating a complex product space and setting technical roadmaps

#LI-Hybrid