Technical Program Manager, HPC

Description

The company builds frontier AI models and products, taking them from pretraining through post-training, evaluation and launch to millions of users. The infrastructure behind this work runs at very large scale, and its reliability, availability and operational readiness keep researchers and engineers productive.

The role is a Technical Program Manager, HPC, responsible for the operational health and reliability of the high-performance computing infrastructure that supports the company's AI work. You will work with HPC engineering, platform, networking, storage, capacity, datacenter and Azure teams to keep production clusters healthy, available and ready for demanding AI workloads.

The job combines HPC infrastructure, production operations, reliability and technical program management. You will set up the mechanisms that catch problems early, speed up recovery when incidents happen, push for permanent fixes, and make sure service commitments are met.

You need enough technical depth to discuss cluster health, compute, networking, storage, schedulers and infrastructure dependencies credibly, along with the program leadership to coordinate complex problems across organizational lines.

Responsibilities

• You will run the operating rhythm and cross-team work needed to keep HPC clusters healthy and reliable.
• You will build a clear picture of fleet and cluster health by combining operational signals, incidents, Nodes-In-Service and Racks-In-Service, infrastructure risks, dependencies and service-level objectives, so teams can focus on the issues hurting production most.
• During critical incidents, you will help coordinate the response across HPC and Azure teams. This means setting clear ownership, escalating quickly, keeping stakeholders informed, and following each case through mitigation, root-cause analysis and corrective actions.
• You will spot recurring reliability patterns across incidents and turn them into lasting engineering and operational fixes. You will also build relationships with Azure HPC, networking, storage, capacity, datacenter and other partner teams so dependencies are understood, escalation paths exist, and SLAs and SLOs are met consistently.
• Own end-to-end programs for HPC production health and operational readiness, improving cluster availability, reliability and service performance.
• Set up routines for cluster health: SLA and SLO reviews, NIS and RIS tracking, incident trends, risks, dependencies and corrective actions.
• Lead critical incident management, covering triage, escalation, mitigation, stakeholder updates, root-cause analysis and follow-through across HPC and Azure teams.
• Find systemic reliability problems in operational data and incident patterns, and work with engineering teams on lasting fixes.
• Build cross-functional partnerships with Azure compute, networking, storage, capacity, datacenter and platform teams to resolve dependencies and set up escalation paths.
• Define and track operational metrics and dashboards that show cluster health, availability, incidents, risks and SLA and SLO performance in a useful way.
• Manage production risks and dependencies actively, raising issues with engineering and leadership before they block production.
• Align engineering and service owners around shared reliability goals and accountable delivery, without direct authority over them.
• Report production health, major incidents, risks and trade-offs to leadership, with clear ownership and action plans.

Requirements

• Significant experience in technical program management, infrastructure engineering, production operations, site reliability or large-scale distributed systems.
• A track record of delivering complex programs that span several engineering organizations.
• Working technical knowledge of large-scale infrastructure, such as HPC, compute, accelerators, networking, storage, schedulers, capacity, telemetry or distributed systems.
• Experience running production incident management, including escalation, mitigation, root-cause analysis and corrective actions.
• Experience working with SLAs, SLOs, availability metrics and production-health indicators.
• Ability to use operational data and incident trends to find systemic problems and drive reliability improvements.
• Strong skill at building relationships and influencing engineers across organizational boundaries.
• Excellent written and verbal communication, able to turn complex technical issues into clear risks, decisions and actions.
• Preferred: experience supporting GPU or accelerator-based HPC or AI infrastructure at scale.
• Preferred: experience with cloud infrastructure across compute, networking, storage, capacity or datacenter operations.
• Preferred: experience setting up reliability programs, operational scorecards, health dashboards or incident-management mechanisms for large infrastructure environments.

Conditions

Position level: Technical Program Management IC5.

Base pay across the U.S. is USD $142,800 to $274,800 per year.

In the San Francisco Bay Area and New York City metropolitan area, base pay is USD $188,000 to $304,200 per year.

The role may be eligible for benefits and other compensation.

The position stays open for at least 5 days, and applications are accepted on an ongoing basis until it is filled.

Benefits

• Certain roles may be eligible for benefits and other compensation.