The Machine Learning Platform, Infrastructure and Systems (MIPS) org at Etsy is responsible for enabling machine learning centrally across all ML practitioner groups. One challenge we typically face is building this centralized tooling in a way that serves the diverse set of Search, Ads, Recs, and Risk use cases. In this session, we’ll talk about our practices for enabling highly reusable ML tooling through the lens of a specific project we recently finished that allows teams to use scalable approximate nearest neighbors search.
