You’re a data scientist, and you’ve got a predictive model — great work! Now what? In many cases, you need to hook it up to some sort of large, complex software product so that users can get access to the predictions. Think of LinkedIn’s People You May Know, which mines your professional graph for unconnected connections, or Hopper’s flight price predictions. Those started out as prototypes on someone’s laptop, and are now running at scale, with many millions of users.
Even if you’re building an internal tool to make a business run better, if you didn’t build the whole app, you’ve got to get the scoring/prediction (as distinct from the fitting/estimation) part of the model connected to a system someone else wrote. In this blog post, I’m going to summarize two methods for doing this that I think are particularly good practices — database mediation and web services.