AI Runtime Engineer

The AI Runtime Engineer at Modular is responsible for owning and optimizing a runtime across various CPU, GPU, and accelerator hardware platforms to achieve state-of-the-art performance for diverse AI models. This role involves designing cross-stack optimizations to address performance issues like CPU overhead and data locality, while collaborating with vendor-specific networking libraries to enhance data transfer speeds. The position requires working closely with internal teams and customers to understand performance requirements and develop automated systems for performance analysis and benchmarking.