Ai Infrastructure Site Operations Managers

About the role

Responsibilities

• Regional GPU Infrastructure Operations: Own the operational health and availability of all NVIDIA DGX B300 / GB300 and HGX B300 / GB300 GPU cluster environments at the regional site — achieving and sustaining ≥99.9% infrastructure uptime
• Serve as the single point of operational accountability for the regional site — escalation owner for all P1 and P2 infrastructure incidents
• Manage and coordinate 24/7 operations coverage across AI Systems, Network & Security, Platform, and Site Resilience engineering teams
• Oversee physical data center operations including power, cooling, physical security, and facility management in coordination with data center facility teams
• Ensure all GPU cluster operations comply with NVIDIA NCP reference architecture operational standards and certification requirements
• Maintain operational readiness for all compute, storage, fabric, and software stack components across the regional portfolio
• Team Leadership & People Management: Lead, manage, and develop a regional team of 8–15 engineers spanning AI Systems, Network & Security, Platform, and Site Resilience disciplines
• Define team structure, shift rotations, on-call rotas, and escalation paths to ensure 24/7 operational coverage with no single point of failure
• Conduct regular 1:1s, team performance reviews, and career development conversations — driving retention and capability growth
• Build a high-performance, collaborative operations culture with a strong focus on reliability, accountability, and continuous improvement
• Partner with VP Technology and HR to recruit, onboard, and develop engineering talent aligned to NVIDIA NCP operational requirements
• Define training and certification plans for the team — ensuring all engineers maintain relevant NVIDIA, vendor, and professional certifications
• Incident Management & Service Continuity: Lead and own P1 and P2 incident response for all regional GPU infrastructure events — ensuring rapid detection, containment, recovery, and communication within defined SLAs
• Chair post-incident reviews and blameless post-mortems — ensuring root cause analysis is completed, findings documented, and remediation actions tracked to closure
• Maintain and exercise incident response runbooks covering GPU node failure, InfiniBand fabric partition, storage cluster loss, power/cooling failure, and security incidents
• Manage the on-call escalation framework ensuring 24/7 P1 response within 15 minutes across all regional engineering teams
• Report incident trends, MTTR performance, and reliability improvements to VP Technology through weekly and monthly operations reviews

Skills

• NVIDIA DGX
• HGX
• InfiniBand
• GPU infrastructure
• data center operations
• incident management
• P1/P2 incident response
• on-call management
• 12/24/7 operations
• team leadership
• service continuity
• root cause analysis
• post-incident reviews
• capacity planning
• vendor certifications