Foresight Institute
- Location: Europe / San Francisco / remote
- Scope: 80-100%
- Starting date: ASAP
About the role
Foresight Institute seeks a Program Manager, Compute to run our global compute program across sites in the US and Europe and our affiliated AI Nodes.
You will track inventory and usage, run procurement and logistics, and manage vendors and partners, so grantees, fellows, and nodes get reliable compute on budget. You coordinate the people who fix problems; you do not write code or repair hardware yourself. The first 6–12 months focus on buildout support before the role moves to steady-state operations.
What you’ll do
Inventory, Usage & Budget
- Maintain one global inventory of compute assets and contracts.
- Track and allocate capacity across sites, projects, grantees, and nodes; flag idle and overloaded resources.
- Manage the annual compute budget and opex tracker. Report costs and usage monthly for finance, funders, and grant reports.
Operations & Incidents
- Monitor site dashboards and alerts. Resolve issues through vendors and contractors, or escalate to engineering.
- Own the on-call rota and escalation paths. Write post-incident summaries and track fixes to closure.
- Schedule maintenance windows, tests, and drills with the engineering team.
Procurement & Logistics
- Run procurement end to end: requirements, quotes, approvals, ordering, and handoff to DevOps.
- Manage shipping, customs, spares, and end-of-life across countries.
- Track lead times and supply risks; keep backup sourcing options.
Vendors & Partners
- Manage contracts with hardware vendors, cloud providers, data centers, contractors, and compute partners.
- Negotiate pricing, credits, and donated compute. Manage SLAs, warranty claims, and invoices.
Coordination & Documentation
- Co-develop the annual compute plan and KPIs with the CEO, COO, Berlin AI Node Manager, and SF Community Manager.
- Onboard grantees and affiliated nodes to Foresight’s compute with node and community leads.
- Set shared standards for access, usage, and security with DevOps. Maintain SOPs, runbooks, and asset records.
Impact
- Link compute usage to outputs (papers, code, demos) and contribute to Foresight’s impact management.
What we’re looking for
- Technical fluency: Reads monitoring data, understands hardware and infrastructure alerts, and represents issues accurately to engineering.
- Experience: 1+ years coordinating IT, data center, lab, or facility operations, ideally multi-site and multi-country.
- Vendor management: Negotiates contracts, manages SLAs, and chases RMAs.
- Process discipline: Uses runbooks, checklists, and incident logs, including under pressure.
- Independent judgment: Comfortable as the sole ops person; knows when to escalate and when to resolve.
- Plus: HPC/GPU cluster or liquid-cooling exposure, cross-border logistics, utility permitting, monitoring/ticketing tools, an additional site-country language.
- Not required: software development, Linux administration, hardware repair, or 24/7 on-site presence.
Don’t hesitate to apply even if you don’t meet every criterion.