Technical Program Manager, Global Infrastructure Planning

OpenAI•San Francisco, CA

12h•$207,000 - $285,000•Hybrid

About The Position

The Technical Program Manager will own CPU capacity planning and management across Research and Compute, leading integrated planning for CPU, storage, and network capacity across OpenAI’s global fleet. The role connects workload demand and GPU deployment plans with the foundational resources needed to support pretraining, research, and application infrastructure at scale. The position partners with workload owners, infrastructure engineering, network deployment teams, Finance, Procurement, and cloud providers to turn forecasts into supply and allocation plans. It requires understanding workload architecture, guiding decisions when capacity is constrained, and improving utilization while reducing waste. Planning spans multiple years and ongoing capacity management across regions and cloud providers, including AWS, GCP, Azure, and OCI. The role is based in San Francisco, California, with a hybrid work model requiring three days in the office per week. Relocation assistance is offered to new employees.

Requirements

At least 5 years of experience in technical program management, capacity planning, capacity engineering, or infrastructure management, including ownership of complex programs at scale.
Direct experience managing demand forecasting, supply planning, allocations, and capacity operations for CPU or cloud compute infrastructure.
Understanding of how forecasts, inventory, quotas, delivery schedules, and utilization inform capacity decisions.
Technical depth in one or more infrastructure domains and ability to reason across dependencies among CPUs, GPUs, storage, and networks.
Understanding of how workload architecture drives CPU and memory needs, storage capacity and throughput, network bandwidth and latency, and regional placement.
Experience working across multiple cloud providers and understanding how provider capabilities, quotas, regional constraints, and commercial commitments affect capacity plans.
Experience building or improving capacity planning tools, data models, or automation, and comfort working directly with data to validate assumptions, model scenarios, and measure outcomes.
Experience delivering measurable infrastructure utilization or cost improvements through reclaims, rightsizing, modernization, workload placement, or other efficiency programs.
Ability to lead decisions across engineering, research, product, Finance, and external providers while balancing technical requirements, delivery risk, and cost.
Clear communication skills with working teams and leadership.
Resourcefulness and ownership in ambiguous environments, with the ability to build simple processes that scale.
Interest in and thoughtful consideration of the impacts of AGI.

Responsibilities

Lead CPU demand planning across Research, Compute, and application infrastructure, maintaining forecasts for workload growth, regions, hardware requirements, timing, and acceptable alternatives.
Translate workload growth and GPU fleet plans into integrated CPU, storage, and network requirements, partnering with engineering on capacity, throughput, bandwidth, connectivity, and placement.
Reconcile current and incoming supply with demand, model constrained scenarios, and drive allocation and prioritization decisions.
Lead CPU allocation planning and fulfillment, including quota transfers, rebalancing, and reclaims.
Partner with network engineering and deployment teams to translate regional demand into bandwidth and connectivity plans, ordering inputs, and delivery milestones.
Coordinate CPU and network allocations as capacity is deployed across the fleet.
Assess workload architecture and performance requirements to challenge assumptions and influence technical decisions about resource sizing, regional placement, data movement, storage choices, and alternatives when supply is constrained.
Drive fleet-efficiency programs to reclaim unused capacity, rightsize workloads, modernize infrastructure, and improve placement or resource sharing; measure released capacity and cost improvements while preserving performance and reliability.
Scale integrated capacity planning across CPU, storage, and network, improving demand intake, forecasts, scenario models, inventory and allocation records, and workflow automation.
Partner with cloud providers, Finance, Procurement, and engineering on capacity orders, commitments, and renewals, evaluating regional availability, delivery lead times, cost, and technical fit and resolving supply gaps.
Align capacity plans with cluster and application deployment milestones and ensure CPU, storage, and network dependencies are ready when workloads need them.
Communicate demand, supply, allocation tradeoffs, forecast changes, and risks to teams and leadership, turning competing needs into decisions, accountable owners, and executable plans.