Join Google Cloud as a Senior Software Engineer, focusing on ML Compilers, Frameworks, and Performance. You will be instrumental in developing cutting-edge ML infrastructure that powers Google's products and services, impacting billions of users worldwide.

This role is for innovative engineers who bring diverse perspectives to build next-generation technologies. You'll work on critical projects within Google Cloud, with opportunities to explore various domains including distributed computing, large-scale systems, AI, and more.

As part of the Core ML organization, you will contribute to the central ML infrastructure and tooling used across all Google product areas. Your work will drive ML excellence and push the boundaries of AI development globally.

Drive continuous enhancements for Google's machine learning software and hardware stack, serving both internal teams and Google Cloud Platform customers. Develop a comprehensive understanding of Google's ML training and serving infrastructure, including frameworks like JAX and PyTorch, Accelerated Linear Algebra (XLA), and the runtime stack.

Identify and implement performance optimizations for ML workloads through in-depth debugging and custom solutions. Collaborate with cross-functional teams to understand performance optimization needs and contribute insights to open-source ML inference frameworks like TorchTPU, vLLM, and SGLang.

Support emerging ML paradigms, such as horizontal scaling for advanced TPU chips, by contributing across the entire stack and performance analysis tools.

Possess a Bachelor's degree or equivalent practical experience, coupled with at least five years of software development expertise in C++ or Python.

Demonstrate practical experience in machine learning (ML) infrastructure development or ML performance engineering. Familiarity with ML compilers, their internal workings, and the ability to develop compiler optimization passes is highly valued.

Experience with accelerator hardware architectures like TPUs and GPUs, alongside expertise in ML inference frameworks (e.g., vLLM, SG Lang, Pathways) and ML frameworks (e.g., TensorFlow, JAX, PyTorch, Keras), is preferred.