Build and validate simulators for AI cluster performance, energy consumption, scheduling, and resource optimization. Benchmark workloads on simulated and real GPU or accelerator hardware, compare results against measurements, and support technical product research. The role requires advanced computer science or electrical and computer engineering knowledge, coding expertise, GPU and accelerator architecture understanding, LLM training and inference knowledge, and familiarity with simulation, benchmarking, power measurement, and cluster resource allocation.

Build end-to-end GPU infrastructure features across cloud backends, node agents, and workload containers. Develop APIs, streaming dashboards, AI-agent validation workflows, observability systems, and automated diagnostics. Debug production GPU, networking, driver, thermal, and hardware failures; profile and optimize LLM inference; implement tracing, metrics, logging, tests, and CI/CD. Work with distributed GPU systems, inference engines, Kubernetes, and multi-vendor environments including NVIDIA, AMD, and Intel.