Delivery Director, Capacity Programs (AI Infrastructure)

Location: San Francisco, hybrid

Salary: $265,000–$285,000 annually, plus equity

Company

Baseten provides applied AI infrastructure, inference systems, and developer tooling that help AI companies bring cutting-edge models into production.

What you will do

• Own end-to-end delivery of on-premises data center builds, including colocation expansions, power and cooling readiness, rack-and-stack, network fabric bring-up, and hardware acceptance testing.
• Drive GPU cloud and neo cloud capacity delivery, including contract milestones, capacity ramps, SLAs, and go-live readiness.
• Build master delivery schedules across concurrent, multi-site, and multi-vendor programs.
• Coordinate engineering, data center operations, procurement, finance, legal, executives, vendors, and infrastructure partners.
• Manage risks, dependencies, delivery governance, capacity tracking, and executive reporting.
• Lead post-mortems and continuous improvement to improve the cost, speed, and predictability of infrastructure builds.

Requirements

• 8+ years of experience in technical program management, delivery management, or infrastructure program leadership, including 3+ years managing large-scale physical infrastructure or data center delivery programs.
• Experience delivering on-premises data center builds from planning through go-live, including power, cooling, networking, and rack-and-stack operations.
• Direct experience with GPU cloud or neo cloud capacity delivery and knowledge of GPU cluster architectures, InfiniBand or RoCE fabrics, rack power density, and liquid cooling.
• Experience managing complex, multi-vendor, multi-workstream programs with hard deadlines and significant financial or business impact.
• Strong executive communication, vendor management, negotiation, risk management, and structured problem-solving skills.
• Ability to operate effectively in an ambiguous, fast-moving, high-growth environment.

Nice to have

• Experience at a hyperscaler, neo cloud provider, colocation company, or AI infrastructure company.
• Familiarity with NVIDIA H100, H200, or GB200-class systems, GPU hardware lifecycles, power and thermal constraints, and compute supply chains.
• PMP, PgMP, or equivalent program management certification.
• Experience building delivery or program management functions from the ground up.

Culture & Benefits

• Competitive compensation with meaningful equity.
• 100% coverage of medical, dental, and vision insurance for employees and dependents.
• Flexible PTO and a company-wide winter break.
• Paid parental leave and a fertility and family-building stipend.
• Company-facilitated 401(k).
• Exposure to ML startups and AI infrastructure projects.