Teams at NVIDIA

Serve as the senior technical contact for OEM Federal deployments of NVIDIA GPU-accelerated platforms. Own customer issues end to end, including troubleshooting, failure analysis, RMA procedures, root-cause investigation, and coordination with OEM support and NVIDIA engineering. Support secure Federal environments such as FedRAMP, IL5/IL6, and air-gapped SCIF facilities while maintaining sanitization and chain-of-custody requirements. Analyze multi-GPU systems, firmware, operating systems, networking, storage, and distributed workloads, with up to 35% travel.

Design and build intelligent compiler systems, compiler-oriented agents, code transformation workflows, and numerical verification infrastructure. Develop systems for low-level IR reasoning, code generation, optimization, differential testing, regression detection, and correctness validation across GPU-centric workloads. Collaborate with compiler, CUDA, runtime, library, and hardware teams to integrate capabilities into NVIDIA products while optimizing performance, scalability, and numerical correctness.

Develop compiler optimization algorithms for deep learning workloads, optimizing JAX and OpenXLA performance on NVIDIA GPUs. Design graph partitioning, tensor sharding, code generation, and performance-tuning solutions for distributed training and inference. Collaborate with deep learning framework and GPU hardware teams, implement user-facing JAX features, and contribute to production-grade software using MLIR, LLVM, and OpenAI Triton.