Technical Program Manager, Research Compute Services
About The Position
The Technical Program Manager will join the engineer-first Research Compute Services team and build and own the tools, data, and platform that translate Research plans and workload signals into compute demand and manage allocations across OpenAI’s compute fleet. This is a hands-on technical role, working alongside researchers and engineers to prototype solutions, automate workflows, and deliver tools used in daily work. The role combines technical judgment and program leadership to connect capacity planning with deployment, scheduling, and optimization systems. The position works across Research workloads, including pretraining and research inference, to ensure evolving requirements shape compute services. The goal is to help researchers express their needs, understand allocations, and use compute effectively as priorities and workloads change. The role also coordinates delivery across Research, Research Compute Services, Capacity Planning, and infrastructure engineering, making progress, constraints, tradeoffs, and decisions clear to teams and leadership.
Requirements
- Strong engineering foundation developed through relevant education or demonstrated technical experience.
- Experience leading and delivering complex technical programs involving large-scale compute, distributed systems, infrastructure platforms, or research environments.
- Practical experience with capacity planning, including supply and demand management, compute allocations, and supporting tools and data.
- Experience personally building or materially contributing to tools, data systems, or automation used by engineering or research teams, with comfort working directly with code and operational data.
- Familiarity with Kubernetes deployments and scheduling systems, including how resource constraints and placement decisions affect workload execution.
- Understanding of optimization solvers and the ability to reason about their use in fleet-wide allocation and local scheduling, including constraints, objectives, and tradeoffs.
- Ability to work closely with researchers to understand changing requirements and translate them into practical, useful systems.
- Ability to own ambiguous technical problems, establish a path forward, and drive complex work through delivery and adoption.
- Ability to communicate technical decisions, constraints, and priorities clearly across engineering, Research, and leadership.
- Resourcefulness and comfort with changing priorities, with motivation to improve researcher productivity and compute efficiency.
- Interest in and thoughtfulness about the impacts of AGI.
Responsibilities
- Work directly with researchers across pretraining, research inference, and other workloads, translating their needs into improvements to compute services, tooling, and scheduling.
- Own the roadmap and delivery of tools, data, and platforms for capturing compute demand, supporting allocation decisions, and managing allocations from requirements and design through adoption and ongoing improvement.
- Develop systems that translate changing research plans and workload requirements into concrete capacity needs, including resource requirements, timing, and constraints.
- Partner with Capacity Planning, Research, and infrastructure teams to reconcile compute demand with available supply and build tools and workflows for scenario evaluation, change management, and visible allocation tradeoffs.
- Create data models, views, and automation that show researcher needs, available capacity, allocations, and compute usage; improve self-service workflows for requesting and managing compute.
- Partner with engineers working on Kubernetes deployments and scheduling systems so allocation plans account for workload constraints and translate into usable resources.
- Work with engineers to evaluate optimization approaches, including fleet-level and local scheduling solvers; help define constraints and objectives, assess tradeoffs, and measure fleet-utilization improvements.
- Use workload and fleet data to identify bottlenecks, unused capacity, and recurring manual work, then build durable improvements to the tools and systems involved.
- Lead technical delivery across Research, Research Compute Services, Capacity Planning, and infrastructure engineering, owning plans, dependencies, decisions, adoption, and operational ownership.
- Communicate demand, allocation constraints, delivery progress, and decisions requiring attention to working teams and leadership.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Job Type
Full-time
Career Level
Mid Level
Education Level
No Education Listed
