When devising your data framework, you should be considering an extensible model that ensures and proves the trustworthiness of infomation you saw fit to publish.  It should be able to integrate new data rapidly and not be held up by time required to enhance the data model or time required to test and productionalize new scripts required to integrate required data.  Both of these thoughts are contrary to the production ready disciplines we have instilled on our client's ability to use data required to capitalize on new opportunities made available in the marketplace.

Consider the components of the integration model presented, that being a process to ensure the integrity of data before being staged, an automated discovery of new rules, and the publication of quality asessments and data lineage statements as well as data. 

Most data integration models in place today are fine for batch publication cycles with discrete windows to review and remediate the quality of small amounts of data published at regular intervals.  This model no longer reflects reality.

Your integration model should include:
      Automated discovery of business rules based on data
      Scoring incoming data to determine its soundness
      An active data model subject to organic growth
      A process to spotlight the most relevant data
      Routine data profiling as part of managing data assets
      Publication of Data Quality & Lineage along with data

If your integration processes do not include such facilities, be prepared to upgrade them.  InfoSight Partners welcomes the opportunity to help you identify and develop remediation plans to shore up any difficiencies.