Quality Lead, YouTube Global Vendor Operations

Minimum qualifications: Bachelor's degree or equivalent practical experience. 5 years of experience identifying process issues or working with cross-functional teams. 5 years of experience with contact center metrics and QA methodologies.

Preferred qualifications: Experience with LLM-based evaluation frameworks and prompt engineering. Experience in stratified sampling to identify specific interactions for human validation, ensuring that human efforts are focused on high-signal or ambiguous cases. Familiarity with QA automation and sampling tools. About the job A problem isn’t truly solved until it’s solved for all. That’s why Googlers build products that help create opportunities for everyone, whether down the street or across the globe. As a Program Manager at Google, you’ll lead complex, multi-disciplinary projects from start to finish — working with stakeholders to plan requirements, manage project schedules, identify risks, and communicate clearly with cross-functional partners across the company. Your projects will often span offices, time zones, and hemispheres. It's your job to coordinate the players and keep them up to date on progress and deadlines. The YouTube Global Vendor Operations team develops practical ways to address some of the most complex business issues to keep YouTube growing. We anticipate how decisions are made, persistently explore and uncover the business needs of our key clients, and understand how our range of product and service offerings can enable their business success.

We lead the global vendor quality program which enables our suppliers to deliver amazing experiences for YouTube’s consumers, creators and partners.

Responsibilities Design Quality Lead (QA) Frameworks to ensure alignment with key business outcomes. Monitor AI QA model performance and accuracy using metrics such as Precision, Recall, and Human Inter-rater Agreement. Manage the ingestion of human-validated Golden Labeling from manual overrides to drive continuous, high-fidelity model retraining. Implement blind review protocols and complex multi-tiered human + AI arbitration to prevent Automation Bias. Design stratified sampling methodologies that prioritize high-risk or ambiguous interactions (e.g., cases where AI confidence is low) for human review. Partner with the training team to remediate quality gaps.