How the pega could get access to the training data easily:  most of our training data comes from Redshift. Do we have to use the AWS to collect data first and then feed the data to the pega and use the built-in model or import the model. If so, that makes not much difference with our current style: predict using AWS and import the result to pega.