Real-Time IntegratedML Utility

Currently, there is no simple way to run IntegratedML predictive models on real-time data as it is ingested into IRIS. The best method we’ve found for performing these real-time predictions is by creating a custom business process that includes a SQL action to run a Predict statement. Ideally, a user would not be required to know the IntegratedML SQL syntax and would be able to use predictions in a variety of places (i.e. Business Rules, Data Transformations, etc.). We are interested in finding a way to seamlessly integrate the IntegratedML prediction capabilities into all the components of the interoperability framework. One idea could be to create a utility function that allows a user to choose a model and property to predict from dropdown lists. This utility should also allow the user to define where/if they want the result of the prediction to be stored, whether in the target message of a data transformation, an existing database table, or somewhere else.