Product Quality Engineer — AI Compute Hardware

Meta; Bellevue (WA); Hybrid; USD 120,000 - 180,000; Full time

Job summary

Meta's Bellevue, WA team is seeking a Product Quality Engineer to drive manufacturing readiness for next‑gen AI/ML hardware. You will collaborate with design engineers and global partners to ensure quality and reliability at hyperscale volumes.

You will lead yield, throughput, and cycle-time targets, develop test requirements from NPI to MP, and perform robust FACA-based analyses while traveling up to 25% to supplier sites and optimizing power and performance during validation.

Responsibilities

• Drive manufacturing metrics (yield, throughput, cycle time, first-pass yield rate) across NPI and MP phases, tracking progress against established targets and escalating risks proactively
• Develop, validate, and implement manufacturing and test requirements for advanced compute and GPU systems from NPI through to mass production (MP)
• Evaluate and enhance test coverage, methodologies, and infrastructure to improve hardware reliability and performance at scale against defined targets
• Drive hardware debugging efforts, contributing to the Major Issue List and executing root cause analysis and Failure Analysis and Corrective Actions (FACA) to resolve hardware issues and meet yield targets
• Travel internationally to supplier sites (up to 25%) to manage hardware validation, ensure compliance with test requirements, and maintain technical partnerships
• Partner with cross-functional engineering teams on power optimizations and performance tuning for compute and GPU systems during manufacturing validation

Job description

Meta's Bellevue, WA team is seeking a Product Quality Engineer to drive manufacturing readiness for next‑gen AI/ML hardware. You will collaborate with design engineers and global partners to ensure quality and reliability at hyperscale volumes.

You will lead yield, throughput, and cycle-time targets, develop test requirements from NPI to MP, and perform robust FACA-based analyses while traveling up to 25% to supplier sites and optimizing power and performance during validation.