One of the fundamental challenges facing today's engineers is to put Big Data assets to work deeply within product experiences. It's not sufficient to run a single insightful report on your data warehouse and declare victory. Today's customers expect a personalized, relevant, and delightful experience in the products they use, and that means using machine learning and other data-focused disciplines directly from within engineering teams.

In this talk, I'll outline some of design patterns that have proven useful in the application of machine learning to high-volume production systems. I will also describe some of the team dynamics that emerge in mixed-discipline teams of scientists and engineers. Finally, I will describe some of the ways eBay has leveraged data science in three diverse and high-impact areas: Search, Marketing, and Trust.