At the 7th International AJAX World RIA Conference & Expo held June 22-23, 2009, in New York, Suresh Chandrasekaran, leader of sales and marketing in North America and APAC for Denodo Technologies, discussed the Enterprise Data mashup platform and how to build virtual data services.

"The basic problem that Data Mashups solve is the integration of a lot of information," noted Chandrasekaran. "The function of the Enterprise Data mashup platform is basically to go across unstructured and structured data and have a common platform to create virtualized data services."

Data Mashups is a holistic approach to data integration that deliver virtualized or batch access to internal and external information as Data Services. It is also significantly faster and less costly but works with existing middleware. Chandrasekaran explored three design patterns for Data Mashups: a) Data Services to support Single View, b) Operational BI & Composite Dashboards and c) Web Automation for Partner Processes

He also discussed the data services layer in SOA architecture, and how the SOA needs to be modularized as well as how technologies have grown up around structured and unstructured web. Most business problems today - customer interactions, BI, compliance, marketing - require access to internal, external, structured and unstructured information. Yet SOA initiatives based on traditional technologies have narrow scope or lack flexibility to solve business needs quickly.

According to Chandrasekaran, BI and SOA are the two big use cases.

Other items he touched on were Web automation and how it's based on example-based heuristics, as well as the reuse factor and flexibility factor of data services.