Technical Program Manager, Compute Infrastructure

The role

This role sits within the compute infrastructure group, which operates a large GPU fleet serving production models and training workloads. The program manager works alongside engineers to bring new capacity online with outside providers, running several parallel bring-up programs and reporting progress up to leadership.

What you would do

• Own delivery of new compute hardware configurations and GPU clusters with external partners, including capacity planning for training and inference
• Plan and track multi-workstream bring-up efforts covering hardware, networking, power, and cooling, including dependencies and critical path
• Liaise with chip vendors to de-risk adopting new hardware platforms across kernel, communications, hardware, and scheduling teams
• Create repeatable program tooling such as roadmaps, milestones, risk logs, and operational runbooks
• Work with engineering to make cluster turn-up more reliable, automated, and faster
• Coordinate physical and logical deployment of network points-of-presence, including on-site installation and rack cabling
• Align security, finance, operations, and product or research stakeholders so capacity ships production-ready
• Resolve handoff gaps between teams and partners and surface systemic bottlenecks with durable fixes

What they asked for

• Hard-science degree or a strong record of engineering work
• 5+ years running programs on major capital or hyperscale infrastructure projects
• Track record of single-point ownership for complex, high-visibility technical delivery under tight deadlines
• Experience managing cross-functional and cross-company teams and keeping decisions and information well organized
• Technically fluent enough to partner credibly with top-tier engineering or research teams
• Experience leading external vendors such as engineering firms, equipment suppliers, or construction contractors
• Ability to design simple, scalable processes and manage logistics or supply-chain dependencies
• Comfortable working through ambiguity in a fast-moving environment

Nice to have

• Interest in and thoughtful views on the societal impact of AGI