Insight garnered through knowledge discovery is usually referred to as an "analytic model." Analytic models help a business predict key business measures. Demand forecasting for predicting sales, customer segmentation for predicting types of customers, and customer attrition for predicting the number of customers who will terminate a relationship with a business, are all examples of analytic models.

Developing analytic models is now a time-consuming process involving highly skilled personnel. The HITC is helping to make this task much more rapid by developing well-understood methodology and technologies that make clear the assumptions made by data-analysis algorithms, and in some cases simplify them.

Once an analytic model has been developed, current technology makes it extremely difficult to deploy the model across an enterprise in a manner that the business can assimilate it. Assimilating a model implies that business users understand the analytic model such that they can appropriately apply it to their business situations. They understand it sufficiently that they may sometimes modify the recommendations of the model or even adapt the model for a changing business. For example, most retail replenishment applications embed an analytic model for demand forecasting. This model is rarely understood by retail buyers, and thus they rarely use it to give a retail business the required accuracy in predicting demand. Research at the HITC is helping to generate insight in a continuous process, helping to transform the analytic model such that it can be easily understood and adapted by business users.