Actuate Embraces Analytic Discovery
In BIRT Analytics, Actuate touts a visual analytic discovery tool.
- By Stephen Swoyer
- February 5, 2013
Talk about a quick turnaround. Late last year, reporting stalwart Actuate Corp. acquired Quiterian, a Barcelona, Spain-based visual analytics specialist.
Just last month, Actuate unveiled BIRT Analytics, a new offering based on the Quiterian technology. Officials concede that this first edition of BIRT Analytics is mostly a repackaging of Quiterian's DDWeb software -- but promise there's more to come.
Quiterian wasn't a start-up; it was founded in 2003. There wasn't much danger of product overlap, either: Actuate already has an OLAP-driven discovery feature (Data Analyzer), which it ships as part of Actuate One, the commercial version of the BIRT open source software (OSS) project. Quiterian's DDWeb technology exposes elements of data mining and predictive analytics in a visual, discovery-like context.
The company liked to position DDWeb as an "agile data mining tool," designed to complement existing business intelligence (BI) and analytic practices.
Call it a case of neatly-dovetailed positioning: Actuate now positions BIRT Analytics as just the ticket to complement the existing BI and analytic practices of BIRT and Actuate One customers.
"We've taken that [DDWeb] product, rebranded it, and done a bunch of other things to it to introduce [it as] BIRT Analytics," says Nobby Akiha, Actuate's senior vice president of marketing. "We categorize it with a name that Quiterian coined called 'visual data mining.' The idea is that it has business user[-class] ease-of-use: [a] user interface for the business user to be able to answer questions that they might have ... [and] do a lot of exploration very quickly."
Exploration or discovery in this context might seem straightforward, but "underlying all of the exploration ... is very sophisticated, advanced analytic, predictive analytic horsepower," Akiha claims.
"Exploration" a la BIRT Analytics sounds a lot like "exploration" a la QlikTech Inc. (developer of QlikView), TIBCO (parent company of Spotfire), or Tableau Software Inc. All three vendors helped popularize the BI or analytic discovery metaphor; at this point, in fact, most big BI vendors market "discovery" themed offerings, too. Most of these -- including QlikView, Spotfire, and Tableau, along with newer solutions from MicroStrategy Inc., SAP BusinessObjects, and SAS Institute Inc. -- likewise embed predictive analytic or data mining capabilities.
Mark Gamble, director of marketing applications for Actuate, concedes as much but claims that BIRT Analytics -- which is powered by a (server-side) columnar analytic repository -- can work with larger data sets than any of these players. (This claim is contested forcefully by users of QlikTech, Spotfire, and Tableau, however.)
"These other tools tend to start to fall down when they start talking about millions of rows of data, or tens of millions of rows; our demo database is over 52 million rows, and [Quiterian] has customers with billions of rows," Gamble points out.
Adds Akiha: "The [BIRT Analytics] repository was based on something built in-house [by Quiterian]. It was built internally at Quiterian from the ground up, taking the concepts and things outlined in several of the academic papers by [Michael] Stonebraker and others."
Actuate Reinvented
Actuate cut its teeth in production reporting, a segment where it's still competitive. Almost a decade ago, Actuate helped kick start BIRT -- i.e., the Business Intelligence Reporting Tool -- an open source software (OSS) reporting project. (Actuate assigned several of its employees to the project.) Over the course of the last decade, BIRT developed into a popular OSS reporting tool; Actuate itself now resells a repackaged and enhanced version of BIRT called Actuate One.
The acquisition of Quiterian and the repackaging of its DDWeb technology as BIRT Analytics give Actuate a discovery-themed offering with creditable analytic chops.
"What we get with BIRT Analytics is the ... [ability] to do clustering, time-series analysis, decision trees, and other, more effective ways of data mining," says Gamble.
"One of the main strengths is the fact that virtually anybody can work against these masses of data and do advanced things like time-series predictions and forecasting."
At this point, integration between BIRT Analytics and Actuate One is comparatively primitive, Akiha concedes. "Although this first instance of BIRT Analytics is still somewhat standalone, we're going to continue to integrate that into the rest of the Actuate One suite," he explains.
"Right now, when you decide that you want to save and share that [i.e., the results of analytic discovery], you can save it to the [BIRT Analytics] repository. Our plans are to extend that out to share it to a traditional BIRT 360 dashboard, and there it would also inherit the characteristics of a traditional BIRT dashboard, including being able to proactively identify a group of users."