Head of Compute Operations

You will manage compute capacity across data-center partners, including hardware, locations, networking, operators, and delivery dates. You will vet new providers, perform diligence and KYC, review hardware specifications, run proofs of concept and acceptance testing, configure access, and coordinate cluster handoffs. You will also create support runbooks, escalate supplier issues, manage RMAs and replacements, and monitor uptime against SLAs.

Responsibilities

• Manage compute capacity across data-center partners
• Onboard and vet compute operators
• Conduct diligence and KYC on new partners
• Review hardware specifications and bills of materials
• Run proofs of concept, acceptance testing, burn-in, NCCL, and interconnect bandwidth validation
• Set up bare-metal and BMC access
• Hand clusters to customers on promised delivery dates
• Create support runbooks and escalate supplier issues
• Manage RMAs and replacements
• Track uptime against SLAs

Requirements

• Experience in datacenter operations, capacity planning, or infrastructure delivery at a GPU cloud, neocloud, hyperscaler, or colocation provider
• Ability to SSH into a node and distinguish InfiniBand from RoCE
• Knowledge of how deposits and prepayment schedules affect cash flow
• Vendor-management experience
• Attention to detail

Benefits

• Meaningful equity
• Health, dental, and vision coverage with 90% coverage for employees and 50% for dependents
• Fitness stipend
• Learning budget