The role involves optimizing distributed machine learning systems to achieve high performance on TPU and GPU clusters. Engineers will work across the entire machine learning stack to improve compute efficiency for both cloud and vehicle environments. Responsibilities include developing new neural model architectures and enhancing training and inference performance. The position requires collaborating with modeling teams to implement compiler optimizations and low-level kernels. Additionally, the role involves integrating state-of-the-art technologies to improve the scalability of machine learning workloads.

Candidates must possess a Bachelor of Science degree in Computer Science or Mathematics, or have at least eight years of equivalent real-world experience. Proficiency in distributed systems design and machine learning efficiency is required, along with solid programming skills in Python and C++. Applicants must have experience with TensorFlow, JAX, and XLA, as well as familiarity with profiling tools. A Master of Science degree in Computer Science or Mathematics is preferred.