PAMELA is an open source modeling tool set implemented in Clojure that was funded by DARPA as a marriage of model-based programming and machine learning. Effective learning has been shown to result in extremely high performance results when done well, but the obstacles are immense from collecting the training data to applying the correct learning methods to integrating back the learning results into the program that needs them. Imagine a model-based program that was integrated with your application that automatically gathered training data, automatically ran the correct machine learning algorithms and automatically inserted the learned results into the program. PAMELA has come a long way since we introduced it in Seattle at Clojure/west because given a plant implemented in Clojure and a model of that program written in PAMELA we can now automate and integrate the whole learning process in exactly that way. In this talk, assisted by a life running demonstration, we will present in detail how the PAMELA tool chain allows in situ learning to happen without having to be an ML expert and without tears.