Principal Applied Scientist

This role is part of the Microsoft Search & Ads Network (MSAN) modeling team, focused on building large-scale machine learning systems for ads retrieval, ranking, user understanding and marketplace optimization across different surfaces. The team develops end-to-end models that predict user engagement and advertiser value - powering candidate generation, relevance scoring, and serving stack ranking that directly impact ad quality, delivery efficiency, and revenue. Responsibilities span the full modeling lifecycle, including training data and labeling strategy, feature and signal design, model development, and rigorous offline and online evaluation. Engineers and applied scientists work closely at the intersection of machine learning, economics, and large-scale systems to deliver high-performance real-time inference and robust experimentation in production.