Technical Program Manager, Compute Qualification

Responsibilities

• Own and continuously improve the end-to-end qualification process for new compute capacity, from initial provider intake through final go/no-go recommendation.
• Run multiple provider evaluations in parallel, setting timelines, tracking status, and keeping every stakeholder aligned on what is needed and by when.
• Partner with infrastructure engineering, network engineering, data center engineering, and SRE teams to plan and coordinate technical validation, then translate their findings into clear decisions for leadership.
• Review provider technical specifications and questionnaire responses for completeness and accuracy, flagging gaps, inconsistencies, and risks that warrant follow-up.
• Conduct first-pass analysis of provider data yourself: compare specifications across suppliers, sanity-check performance claims, and surface issues before deeper engineering review.
• Maintain the standards, templates, and documentation that define what meets spec across compute, networking, storage, power, cooling, and operational support.
• Build a structured, auditable record of evaluation outcomes that informs sourcing decisions and scales the qualification function as the team grows.
• Deep-dive into critical hardware performance metrics, proactively identifying potential bottlenecks in cluster architecture before they impact our end customers training or inference workloads.
• Conduct diligence and work with engineering teams to make assessments regarding technical and operational resilience.

Requirements

• 5+ years in technical program or project management, infrastructure program management, or a comparable technical operations role, ideally involving hardware, data center, or large-scale compute environments.
• Proven ability to run multiple complex, cross-functional workstreams to deadline, with strong organization and stakeholder management.
• Working technical fluency across data center infrastructure: server and GPU hardware, high-performance networking (InfiniBand or Ethernet fabrics), storage, and power and cooling fundamentals; enough depth to read a detailed technical specification and know what to question.
• Hands-on comfort with data: able to write scripts or queries (for example, Python or SQL) to compare, validate, and analyze provider specifications and test results independently.
• Excellent written and verbal communication; able to turn dense technical detail into clear recommendations for both engineers and executives.
• Willingness to travel to provider and data center sites as needed.

Nice to Have

• Experience qualifying, commissioning, or accepting GPU clusters or HPC infrastructure against defined performance and reliability standards.
• Familiarity with AI training and inference infrastructure, including interconnect topologies, cluster bring-up, and acceptance testing.
• Experience in AI/HPC cluster design.
• Background working directly with hardware vendors, colocation providers, or cloud capacity providers.