Senior Software Engineer, Infrastructure

The Planner Infrastructure team at Waymo builds software that runs on the self-driving car, in addition to tools & benchmarking for ensuring that our system is performant and reliable. We are looking for infrastructure engineers to lead and help us take our system to the next level. This will ensure the self-driving car can operate in environments like freeways and crowded city streets.

In this role you'll:

• Report into our Planner Infrastructure TLM
• Ensure that our software is architected for reliability and implemented using advanced technologies and engineering best practices.
• Conduct software testing and validation, including the use of simulation and static analyzers to ensure that there are no reliability problems in our software that runs on the road.
• Ensure that our system can produce safe driving plans in an allotted timing and compute budget (potentially for hundreds of people and cars around the self-driving car).
• Be able to react and change direction at super-human latencies when there are new hazards detected by our perception system.
• Ensure our machine learning models are performing well on GPU/TPU and other hardware accelerators.

At a minimum we'd like you to have:

• BS in Computer Science, Electrical Engineering, Robotics, Physics, Math, or related field (or equivalent experience)
• 4 years of industry experience designing and writing production code
• Proficient in C++, Java, Rust, or other modern languages.
• Experience leading technical contributions and excellent engineering execution.

It's preferred if you have:

• Experience designing, building, or maintaining FAANG-scale infra platforms.
• Experience building performant and easy to use scalable systems and tools.
• Expertise in C++ and C++ specific low-level code optimizations
• Experience in robotics or self-driving cars.
• Experience with ML model inference performance and frameworks like CUDA and JAX.
• Experience designing or writing software for high-reliability systems, performance optimization, or productionizing deep learning models on GPU for low inference latency.

#LI-Hybrid