LLM Inference Engineer

The LLM Inference Engineer role involves owning the end-to-end LLM serving stack on custom hardware, architecting components from serving APIs to KV cache management, and optimizing frameworks like vLLM and SGLang for peak performance. The position requires porting leading frameworks to proprietary accelerators, hunting down bottlenecks, and providing insights to steer future kernel, compiler, and hardware development. This high-impact role bridges cutting-edge custom silicon and production-grade AI to shatter traditional GPU limits on throughput and context lengths.