Exposure to GPU infrastructure and AI workloads, such as NVIDIA DGX and SuperPOD reference architectures, GPUDirect Storage, inference serving frameworks (vLLM, NVIDIA Dynamo, llm-d) and the storage patterns behind training, inference and retrieval is a plus. The CSE owns the technical strategy for each account, earns the trust of architects, platform owners and technology executives, and orchestrates product specialists, professional services and partners to win multi-year, multi-million dollar platform decisions.