Compute Deployment Lead

Lead Compute Deployment

Responsibilities

• Lead compute deployment across sites: the team and process taking GPU nodes from dock to validated clusters.
• Own the compute deployment schedule per hall: thousands of nodes racked, imaged, burned in, and validated in weeks.
• Build the bring-up factory: automation and crew workflows that scale across parallel sites.
• Hold the validation bar: no cluster hands to production without proving its performance.
• Turn empty data halls into serving GPUs in weeks. Deployment owns everything between construction handover and customer-ready compute: racks, fabric, and validation at hundreds of megawatts per site.
• Stand up gigawatts a year, hundreds of megawatts at a time. Data halls go from construction handover to serving GPUs in weeks, across multiple sites at once.
• Deploy 10GW with software, not headcount. Bring-up, validation, and handoff run on tooling the team builds, so the hundredth data hall deploys faster than the first instead of needing a hundred times the people.

Requirements

• Led compute or server deployment at data center scale.
• Built bring-up processes that cut deployment time by a large factor.
• Managed mixed teams of engineers and technicians.
• Refused to hand over hardware that didn't pass, under real pressure.

Nice-to-Haves

• GPU clusters.
• Deployment automation.
• HPC acceptance testing.
• Vendor integration programs.