Software Engineer, Training, ML Infrastructure

at  Waymo

Waymo is an autonomous driving technology company with the mission to be the most trusted driver. Since its start as the Google Self-Driving Car Project in 2009, Waymo has focused on building the Waymo Driver—The World’s Most Experienced Driver™—to improve access to mobility while saving thousands of lives now lost to traffic crashes. The Waymo Driver powers Waymo One, a fully autonomous ride-hailing service, and can also be applied to a range of vehicle platforms and product use cases. The Waymo Driver has provided over one million rider-only trips, enabled by its experience autonomously driving tens of millions of miles on public roads and tens of billions in simulation across 13+ U.S. states.
The Waymo ML Infrastructure team works with Research and Production teams to develop models in Perception and Planning that are core to our autonomous driving software. We ensure our partners by offering the best solutions for the entire model development lifecycle. These solutions are developed in close collaboration with teams at Google. They are geared towards both scaling models and solving problems unique to ML for autonomous driving.
We develop a set of libraries and tools that enhance TensorFlow and JAX, and address scalability, reliability, and performance challenges faced by Waymo’s ML practitioners: training fast and at scale, increasing ML accelerator efficiency, fine-tuning multimodal LLMs for autonomous driving tasks, discovering hyper-parameters, retraining neural networks, computing reliable and noiseless metrics on validation sets, and validating newly trained DNNs when deployed into the full onboard software stack.
We are looking for an individual contributor (IC) who enjoys the following responsibilities, and has the required qualifications listed below:

• Develop the infrastructure components necessary for distributed training, including job scheduling, resource management, data distribution, and model synchronization.
• Implement automation solutions for provisioning, deployment, monitoring, and scaling of distributed training infrastructure to improve operations and reliability.
• Monitor system health, diagnose and troubleshoot issues, and perform routine maintenance tasks to ensure the reliability of the distributed training infrastructure.
• Identify performance bottlenecks and optimization opportunities
• Improve the developer experience and performance of our scalable ML framework