Experience with Azure, Microsoft Foundry or comparable AI gateways and model routers, NVIDIA AI Enterprise and GPU infrastructure, vLLM/NIM or comparable open-weight model serving, Snowflake AI capabilities, MLflow, Kubernetes, MCP (Model Context Protocol), or comparable technologies is preferred. Working knowledge of model serving, RAG, vector databases/search, LLM gateways and model routing, agent frameworks and tool/MCP integration, evaluation, LLM observability (traces, tokens, latency, cost per task), security, and AI workload economics.