Staff Applied ML Software Engineer, Recommendations, YouTube
San Bruno, CA, USA · Staff · Full-time · Posted Oct 21, 2026
- Experience
- 8 years in software development; 5 years building and deploying production recommendation models and 5 years with ML design and infrastructure; 3 years in Applied Machine Learning. Publications in ML or a related field required.
- Education
- Bachelor's degree in Computer Science or a related field, or equivalent practical experience.
About the job
Leads the design and development of YouTube recommendation systems, models, and ML infrastructure to improve video retrieval and ranking.
What you'll do
- Provide technical leadership on high-impact projects and manage priorities, deadlines, and deliverables.
- Lead recommendation system design and implementation, optimize ML infrastructure, and guide model architecture.
- Align and coach distributed teams, and analyze data and logs to improve the product using ML.
What they ask for
- Bachelor's degree in Computer Science or a related field, or equivalent practical experience.
- 8 years of experience in software development.
- 5 years of experience building and deploying production recommendation models, plus experience building architecture across modeling domains; 5 years of experience with ML design and infrastructure; and 3 years of experience in Applied Machine Learning.
- Experience with publications in ML or a related field.
Nice to have
- Master's degree or PhD in Computer Science or a related field.
- 3 years of experience in technical leadership and 3 years working in a complex, matrixed organization on cross-functional or cross-business projects.
Pay
$207000 - $300000 (USD) base salary
Tools
ML, recommendation systems, retrieval, prediction, ranking, embedding, model deployment, data processing
AI focus
Applies machine learning to retrieve and rank videos, develop recommendation models, and improve recommendations; work includes ML algorithms, infrastructure, and model architecture.
Model-written summary · the original posting is authoritative