Data Center Incident Program Manager

Data Center Incident Program Manager | OpenAI

Compute - Remote - US

OpenAI, in close collaboration with our capital partners, is embarking on a journey to build the world’s most advanced AI infrastructure ecosystem. Our Stargate program develops and deploys massive, state‑of‑the‑art data center campuses in partnership with industry leaders such as Oracle today—and through future OpenAI infrastructure projects tomorrow. We design for scale, speed, and reliability, and we need experienced hardware professionals who can help ensure our high‑density compute environment operates at peak performance.

The Data Center Incident Program Manager is responsible for designing, operating, and continuously improving the end‑to‑end incident management lifecycle across mission‑critical data center environments. This role owns the “before, during, and after” mechanics of incidents — establishing standards and playbooks in steady state, serving as (or designating) Incident Commander during active events, and driving structured post‑incident review and corrective action to closure.

The ideal candidate brings operational credibility in hyperscale or mission‑critical infrastructure, demonstrates calm leadership during high‑pressure events, and has a strong bias toward structured documentation, process clarity, and measurable improvement.

• Define and maintain incident severity levels (SEV definitions), classification criteria, and escalation thresholds.
• Establish end‑to‑end incident response standards: protocols, lifecycle stages (declare → stabilize → mitigate → recover → close), and operating cadence.
• Build and maintain governance artifacts: runbooks, war room formats, reporting templates, and decision/communication standards.
• Create and operationalize notification trees, stakeholder comms templates (initial, periodic updates, recovery/closure), and executive escalation criteria.
• Define clear RACI across Facilities, Hardware Ops, Network, Security, and vendor/partner teams, including handoffs and accountability paths.
• Set and manage SLAs/OLAs for acknowledgment, escalation, containment, mitigation, and reporting.
• Implement and run incident management tooling (ticketing, paging, logging) and ensure integrations with monitoring and workflow systems.
• Establish dashboards and program health metrics to track incident performance and readiness.
• Lead readiness activities: tabletop exercises, cross‑functional simulations, IC/Deputy training, and a rotating on‑call IC bench with certification standards.
• Serve as Incident Commander as needed: declare severity, stand up the war room, assign functional leads, and drive structured execution under pressure.
• Maintain real‑time documentation (decisions, timelines, impact scope) and ensure clear restoration objectives and scope control during active events.
• Run post‑incident reviews (PIRs), validate timelines, drive structured RCA (e.g., 5 Whys, Fault Tree), and separate root cause vs contributing factors.
• Define corrective/preventative actions (CAPAs), assign accountable owners, track to verified closure, and elevate overdue actions.
• Publish trend reporting (incident taxonomy, counts by severity, MTTA/MTTR, repeat failure domains) and feed systemic gaps back into design and operations teams.

• 7+ years in mission‑critical infrastructure, data center operations, or reliability engineering
• Direct experience leading major incidents (P1/P0 equivalent)
• Strong familiarity with facilities systems, hardware operations, or network infrastructure
• Demonstrated experience running war rooms and executive updates
• Experience conducting root cause analysis and corrective action tracking
• Ability to remain calm and decisive under high‑pressure conditions

• Experience in hyperscale or high‑density AI compute environments
• Background in facilities commissioning, facility operations, hardware operations, or network reliability
• Familiarity with ISO‑based quality systems or structured operational documentation frameworks
• Experience implementing incident tooling (Pager Duty, Service Now, Jira, etc.)

OpenAI is an AI…