Risk applications provide critical infrastructure for financial institutions. The core of a risk application is an analytic model. Creating and deploying effective analytic models depends on having consistent, accurate and timely data.

In this talk, we describe a service oriented approach to working with data, metadata, and derived data. We also describe a service oriented approach for creating PMML-based analytic models using this data and derived data.

Finally, we describe a framework for measuring the accuracy and consistency of the data, derived data, and analytic models in the framework.