Senior Inference Engineer, GPU Kernel Optimization

Drive performance to the ceiling for LLM inference by building silicon-measured GPU kernel microbenchmarking, end-to-end model performance analysis, and agentic kernel optimization systems. Collaborate across compiler, kernel, hardware, and framework teams to attribute bottlenecks, generate optimization policies, and validate improvements with rigorous CUPTI/NSYS/NCU profiling for production-grade throughput and latency gains.