Consultant – Assurance & Validation, GPU Infrastructure

About the role

The Consultant – Assurance & Validation will provide independent technical assurance and validation of GPU implementations and ongoing operations, acting as the client’s technical conscience to ensure quality, performance, and readiness of the delivered GPU platform.

Key responsibilities

• Develop and execute comprehensive technical acceptance and validation plans covering hardware, fabric, storage, management tools, and performance baselines.
• Independently review and challenge OEM designs, configurations, test results, operational procedures, and documentation.
• Validate critical technical areas such as NVIDIA B300 GPU health, 800G fabric performance, PowerScale/NVMe behaviour, OpenManage/iDRAC, and Fabric Manager effectiveness.
• Supervise provisioning, imaging, and lifecycle management of GPU nodes (HGX servers, driver/CUDA stack, firmware, NVLink/NVSwitch).
• Lead cluster bring‑up, expansion, upgrades, and hardware qualification, and oversee knowledge transfer and operational readiness.
• Conduct root‑cause analysis on significant incidents and ensure corrective and preventive actions.
• Maintain technical evidence packs for acceptance, audit, and compliance, and advise on technical risks and trade‑offs.

Required profile

• 10+ years of experience in GPU, HPC, or large‑scale AI infrastructure engineering with proven validation or assurance work.
• Strong practical knowledge of NVIDIA GPU systems, high‑speed interconnects, enterprise storage, and server management tools (Dell ecosystem preferred).
• Experience in an oversight or assurance capacity rather than pure hands‑on build/operate roles.
• Relevant certifications such as NVIDIA‑Certified Professional (AI Infrastructure, AI Operations, AI Networking, AI Rack and Interconnect), Red Hat Certified Specialist in OpenShift, Dell Technologies certifications, ITIL 4 Foundation, or Linux certifications (RHCSA).

Required skills

• NVIDIA B300 GPU health monitoring
• 800 Gbps fabric performance analysis
• PowerScale/NVMe storage behaviour
• OpenManage/iDRAC and Fabric Manager administration
• HGX server provisioning, CUDA driver stack, firmware, NVLink/NVSwitch
• Cluster bring‑up, expansion, upgrade processes
• Root‑cause analysis and corrective action planning
• Red Hat OpenShift Container Platform 4.6
• Kubernetes orchestration
• Dell PowerEdge hardware and AI solution management