The Decision Optimizer application[1] is an optimization software leveraging various optimization algorithms with the goal to empower non-OR professionals with a tool that seamlessly performs optimal assignment of treatments to a portfolio of customers. The underlying problems are based on the Generalized Assignment Problem. They consider different kinds of constraints (budget, ratio, ...) and allow for generating optimal decision trees. 

The optimization models solved by the application depend on structural inputs (values are certain, like the composition of the portfolio or customer attributes) and uncertain inputs (values are uncertain but predictable, like the impact of treatments on customers). While the tool allows for using and editing the predictive models that are required to produce the uncertain data, the skills required to create these Causal Models[2] is usually not available to the typical Business Analyst. 

To simplify the experience of the user, we developed a component which can be used by Business Analysts with either advanced or basic knowledge of predictive modeling to seamlessly create Predictive Causal Models that are suitable for predicting the impact of the treatment assignment on every customer of the portfolio. The approach is generic enough to be applied to a wide variety of business problems that Decision Optimizer aims to solve.