Architect and deploy with NVIDIA platform tools including Base Command Manager (BCM), NGC, NCCL, NVLink, and CUDA along with LLM inference engines (TensorRT-LLM), production serving frameworks (vLLM, SGLang), inference orchestration (Triton Inference Server, NVIDIA Dynamo, llm-d), and GPU benchmarking and validation tools (MLPerf, NCCL tests, fio, iperf) to deploy, tune, profile, and validate AI cluster performance across compute and networking layers including multi-node training and inference workloads. Deploy, configure, and manage XPU-based clusters (GPU, DPU, LPU, CPU) across bare-metal and containerized environments using workload schedulers (Slurm, Run:ai), Kubernetes orchestration, and container platforms to deliver scalable AI infrastructure services including Bare-Metal-aaS, GPUaaS, AIaaS, Token-aaS, model serving, and agentic AI frameworks.