This presentation will describe a three-year journey to re-platform the core transactional systems of a $2.5 billion company using Service Oriented Architecture (SOA) as the data integration layer, and the contributions the Data Management team was able to make and techniques we used along with lessons learned.

SOA is delivering on its promise of integrating loosely coupled distributed systems using prescribed interfaces, and rapidly developing new business functionality from shared services. SOA initiatives tend to focus on the software development aspects. But the real core of a SOA ecosystem is the data, and representation of the data in an application-neutral way (or canonical form) to facilitate its movement between services and systems, usually as XML. SOA moves data integration challenges, historically encountered in data warehousing, upstream to the operational realm. This makes all the DM disciplines that grew out of warehouse experience enormously relevant to SOA projects, yet too often DM is unfortunately and irrationally sidelined. If you have mature DM practices in your company this need not be the case.

The session will include:

• a brief primer on key SOA concepts and methods
• some of the assets DM can bring to the SOA table and use as levers to be integral participants
• how SOA governance and data governance intersect
• the challenges of integrating XML schema documentation into an RDBMS managed metadata environment
• and more of our specific experiences, with successes and frustrations given equal treatment

Participants should gain an understanding of Data Management's role in the SOA world and some specific notions of how to insert yourself into SOA projects to help make them successful.