This workflow is part of the model factory eco system. Please refer to our white paper for further details. It can be used to do a learning of the thresholod which decides when a model is retrained.
This is a companion discussion topic for the original entry at https://kni.me/w/5lQBt-KYXV5d_9a6