We show how to deal with large data sets, how to clean the data effectively by identifying and removing outliers, and how to validate that the solution conforms to user-defined performance targets. And finally, we explain how the regression solutions can be deployed in production.