This topic outlines the implementation of the control decision model concept into the Decision Server runtime and Decision Manager.
The idea of a control decision model is to allow a succession of decision models to be executed one after another, with the specific models that will be executed, and the order in which they will be executed to be determined by the control decision model.
The control decision model itself is no different to a regular decision model in that rules developers will still write their rules for it within the Decision Manager using the same interface as for regular decision models, except that there will be one mandatory fixed schema that the Decision Server will use to determine which decision model it needs to execute next.
The control decision model itself is no different to a regular decision model in that decision designers will still write their rules for it within the Decision Manager using the same interface as for regular decision models, except that there will be one mandatory fixed schema that the Decision Server will use to determine which decision model it needs to execute next.
A common scenario for this feature is that business policy as defined at a point in time (i.e. effective dated business policy) is captured in 'specialist' decision models, and that these decision models are orchestrated to apply all of the policy that is relevant to the transaction. The role of the control decision model to manage the conditional execution of these specialist decision models. This keeps the control decision model relatively compact, removing specific detailed business logic from the control decision model and placing it in smaller separate units (the 'specialist' decision models).
Updates to the logic of specific policy and its supporting decisions can now be managed with lower risk, as change is only necessary within the specific specialist decision model, and not in a larger more general decision model where a change can have wider implications.