We Are:
The Global AI Infrastructure team is at the center of enabling infrastructure reinvention for the next era of digital solutions powered by AI, accelerated computing, and high-performance workloads. We bring together deep technical expertise across cloud, on-premises, and hybrid environments to design, build, and operate advanced infrastructure that powers AI platforms, GPU-accelerated workloads, large-scale models, simulations, and emerging agentic AI solutions at scale. Our solutions enable some of our most strategic and mission-critical clients to unlock new levels of performance, efficiency, governance, and innovation. Our remit spans the full lifecycle-from strategy and architecture through implementation and operations-driving modernization across the entire infrastructure stack. We collaborate across the ecosystem to harness emerging technologies, fuel growth, and transform industries. In this rapidly growing market, our team is leading the way in shaping how enterprises leverage AI infrastructure to drive breakthrough innovation and reimagine what is possible.
Key Responsibilities:
Design and implement AI infrastructure and accelerated computing solutions, aligning system architecture and deployment roadmaps to industry-specific performance, scalability, resiliency, and governance needs
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
Integrate AI infrastructure platforms with existing IT systems, data pipelines, security frameworks, model-serving endpoints, and enterprise governance controls
Design and implement agentic AI infrastructure by integrating platform services, model endpoints, tool and function calling, retrieval patterns, and workflow orchestration with observability, identity, and policy controls through secure, deterministic APIs to support governed enterprise use cases
Build and integrate MCP servers, tools, connectors, and adapters that allows agents to monitor, troubleshoot, and tune infrastructure to ensure high availability, low-latency networking, and workload resiliency
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
Develop and maintain documentation including architecture diagrams, configuration baselines, and operational runbooks
Provide technical guidance, troubleshooting, and optimization across AI workloads including large-scale training, inference, multi-node simulations, and agentic pipelines while leveraging digital twins to validate infrastructure and drive performance, scalability, energy efficiency, and token cost optimization
Travel may be required for this role. The amount of travel will vary from 25% to 100% depending on business need and client requirements.