Is Data the Missing Last Mile for SOA Success?

Ash Parikh, who manages and blogs about Informatica's SOA and real-time data integration strategy, says data is SOA's last mile. As he explained to Loraine Lawson, companies invested in SOA tools, but forgot to factor in access to quality data. He explains how Informatica's data integration platform can, for example, deliver accurate and timely data to “Web-based, self-service architectures” to address their data-centric problems.

Lawson: You've written and spoken about a “Web-based, self-service architecture built with SOA tools.” What is that? What does that mean?
Parikh: Basically it is a portal and it is exposing that portal for using various profiles and login mechanisms and providing customer information, customer account information, in this case, being a financial services company, to the end user.

I’m not going to name the customer, but basically they're in the financial services space. They were trying to move to a more agile architecture and, at the end of the day, their business interest was to increase the level of customer satisfaction by providing more seamless and advanced or modern customer experience.

If you look at the financial space, they're basically gunning each other down just to get more of the customer’s time. So this pretty good customer had come to us a year and a half ago, saying, “Listen, we are in aggressive acquisition mode. We’re trying to acquire all of these systems and all of these companies and these companies are distributed across the world. A big problem is a single view of a customer is extremely tough to create and is basically a challenge and we’re looking for the right kind of platform to help us do that, as well as a technology or a platform that would complement existing architecture investments that we’ve already made.”

They had huge duplication of customer records, customer data. They didn’t have a single view of that information so (there were) data quality problems, as well as serving up that information in a very agile manner.

Basically, they had all the makings of a service-oriented architecture: They had an ESB, they had a BPM (business process management) product, I believe they were looking into an SOA registry and repository.

They were trying to deal with the data problem in the good old way that people normally do, you know, hand coding, point to point integration, etc. Probably even using some EAI (enterprise application integration) tools to deal with information. However, the information was large in volume and basically had huge data accuracy problems.

Previous Page Next Page