About the job
Discover is a proactive, recommendation-driven surface. We innovate on the Machine Learning (ML) models and applicant generation strategies that filter millions of potential ads down to a high-quality subset. Our work directly determines the upper bound of the entire ads system’s performance. In this role, you will lead the retrieval strategy that connects users with the most relevant ads in their discover feed, a primary driver of Google’s social ads effort. You will retrieve high-quality relevant ads that provide real value to a user.
Responsibilities
Be the primary technical authority for our retrieval stack and move beyond incremental tuning to drive step-functions in ad relevance and quality.
Lead a team of ML engineers to explore the frontiers of embedding-based retrieval, deep learning architectures, and multi-objective optimization (user value vs. advertiser success).
Define the roadmap for retrieval quality. This includes moving from legacy retrieval methods (e.g., two-tower models, transformer-based embeddings, and generative retrieval).
Innovate on how we measure and improve relevance.
Lead efforts to align retrieval outputs with long-term user satisfaction and advertiser Return on Investment (ROI) ensuring we aren't just retrieving clicks, but value.
Design next-generation applicant generation strategies.
Oversee the integration of fresh signals (user intent, content semantics, and social trends) into our retrieval models to capture the dynamic nature of the discover feed.
Qualifications
Minimum
Bachelor’s degree or equivalent practical experience.
8 years of experience with software development in one or more programming languages.
7 years of experience testing, maintaining or launching software products, and 1 year of experience with software design and architecture.
7 years of experience with AI/ML algorithms and tools, deep learning and AI/ML modeling.
Experience building and deploying recommendation systems models (e.g., retrieval, prediction, ranking, personalization, search quality, embedding) in production.
Preferred
Master degree or PhD in Computer Science, or a related field.
8 years of experience with data structures and algorithms and large language model applications.
5 years of experience in a technical leadership role leading project teams and setting technical direction.
Experience developing algorithms (e.g., ranking, recommendations, prediction, search quality, personalization) and software that generates suggestions based on various input and output goals.
Experience in quality working on content recommendation or user personalization problems.
Experience in successfully applying machine learning approaches to real-world content problems, and excellent programming skills in Python.