Posted by: RobBarry
Data infrastructure company Greenplum this month announced a major step toward its enterprise data cloud vision, the Chorus platform. The offering provides cloud-oriented features including self-service provisioning, data collaboration and data services. By plugging an existing datacenter into Chorus, the company says an enterprise can get new insight from previously disparate data silos.
“All these analysts inside an organization – smart guys who understand how to use data – they see the data they want to use for answering business questions spread apart across the organization,” said Ben Werther, director of product management at Greenplum. “And there’s no real way to bring it together and to work together to tackle the business problems that the [data warehouse] isn’t designed to face.”
In essence, Chorus can go out and collect any of the data living in an enterprise infrastructure and make it available to business analysts in a centralized repository with some power user tools. The Greenplum Database that this sits on contains tools for analytics, writing MapReduce programs, matrix math, multilinear regression and so on.
Perhaps the most useful aspect of Chorus is that it lets business analysts find any data that can answer questions that the executives might have about business metrics. Much of this data would otherwise live in disparate silos or “under the hood” of a data warehouse. Now the folks at Greenplum have developed a collaborative platform where data relationships can be analyzed and manipulated far more easily.
This is not to be confused with master data management (MDM), which serves to create a single living set of cleansed data for more traditional enterprise data warehouses. Werther said Greenplum customers have taken up a more agile philosophy and that the notion of a “single version of the truth is kind of a fallacy.”
“It’s about getting all the data in as quickly as you can, in whatever form, and putting it in the hands of analysts who understand where it came from and what level of trust they should associate with it,” Werther said. “But you don’t have to wait a year or six months of process just to get it in their hands.”
Given, MDM systems are often updated far more frequently than every six months – and in more advanced SOAs they may even be updated in near real time. Still, it may be quite useful to have a collaborative tool with a friendly UI for discovering and analyzing enterprise data without getting in the way of the data path. Also appealing, its private-cloud approach means you only need commodity hardware to run it.