Boston-based Epistemic is rethinking the way OLAP is used and essentially bringing the analytics process to the data across a variety of data sources, rather than bringing the data into the process.
With Version 2.0 of the Java-based Epistemic Analytics Toolkit, due out in July, the company is supporting Web services to further its cause: helping companies reach the next phase of enterprise applications, such as ERP (enterprise resource planning), CRM (customer relationship management), and SCM (supply-chain management), according to Vidur Dhanda, founder of the company.
"Analytics is the second phase of enterprise software. The first phase is getting the data in the applications. Now that we've got the data, the question becomes what to do with it," Dhanda said. "There is a value to bringing together the ERP and CRM with the financials."
Web services protocols SOAP (Simple Object Access Protocol) and WSDL (Web Services Description Language), Dhanda said, enable Epistemic to impart analytics on data sources.
"The Web services interface essentially provides an analytics utility as a true service. Given that our toolkit is datasource agnostic, we can inject analytics into a workflow or business process," Dhanda said.
That allows manual processes to be more automated than they currently are, Dhanda added. For instance, users can put rules into an OLAP engine, and then look at a multi-dimensional view.
The toolkit also includes what Dhanda called Dynamic OLAP, or the ability for users to define metrics and rules on the fly. So if a user wants to know who is buying yellow socks, where, and at which store types, that analysis can be run. And when the user then wants to alter the analysis to find out which buyers of yellow socks also buy blue socks, that can be done as well.
"You don't have to bring the data to the process, you can bring the process to the data," Dhanda said.
He added that the advantages include faster analytics, and more accurate results because data can be pulled from a variety of sources, be it enterprise applications or databases.
"The goal is to project the model onto the data, thereby reducing latency, that being the time it takes to run the analytics," said Bob Moran, the vice president and managing director of data knowledge and analytics, at Aberdeen Group, a consultancy in Boston.
Moran continued, however, that Epistemic's approach to solving the latency problem is only one of several, and that companies can also use caching or delta management, which is the practice of sending off only changes to data.
Neither is Epistemic the only analytics company to support Web services. Business Objects, in San Jose, Calif., earlier this year announced a Web services SDK (software development kit) designed to extend the reach of business intelligence applications and extranets as Web services, which will be available later this month.
In addition to the Web services support in Epistemic Analytics Toolkit, Version 2.0 includes performance gains and tweaks on the information delivery end that make it easier to develop Web-based applications without bothering with HTML coding, Dhanda said.
Epistemic takes its name from the adjective's definition: of or related to knowledge or knowing.