RealTime IT News

Open Source Business Intelligence Gets 'Cloudy'

Business intelligence (BI) has traditionally meant on-site deployment and direct connections to data warehouse technologies.

Things are changing, though with the release of Jaspersoft BI Suite 3.5, which will enable users to run their analysis without the need to always have a data warehouse or even a local on-premises installation.

Jaspersoft's open source BI competitor Pentaho is also improving its BI suite, with a new release targeted at the cloud.

Together, the open source BI vendors are trying to shake off the notion that open source merely replicates what already exists in proprietary software. Their new BI releases come at time when enterprises are increasingly looking to cut costs and, in many cases, are exploring Software as a Service (SaaS) and the cloud as vehicles of cost optimization.

"I think we've done a lot of clever thinking internally to help us understand what we need to do differently," Brian Gentile, CEO of Jaspersoft, told InternetNews.com. "We know that as an open source company, if all we do is replicate the BI stacks of the past, we will surely have missed a big opportunity. So I think version 3.5 is a perfect example of doing it better."

With Jaspersoft 3.5, a key new feature is the ability to do integrated multi-dimensional analysis in memory. Gentile noted that typically, the heavy lifting of BI analysis could only be done with a backend data warehouse and OLAP . That's not to say that Jaspersoft 3.5 now eliminates the need for OLAP altogether.

"It's a matter of segmentation," Gentile said. "We believe fully that the data warehouse and OLAP markets are alive and well. Bringing the power of a data warehouse down to a broader class of user is our goal. The fact is that more businesses today need access to more data with more analytics skills."

Sherman Wood, director of BI at Jaspersoft, explained that the company compressed results in memory to allow large data sets to be pulled in. He claimed that with Jaspersoft 3.5, a user could run a 20-million-row dataset in memory with 6 to 8 GB of RAM on a server. As a result, Jaspersoft 3.5 could handle datasets that contain gigabytes of data. Wood added that where OLAP is still needed is when a user is pulling in terabytes (TB) of data.

SaaS deployment

Another key feature of Jaspersoft 3.5 is its support for SaaS deployments. Gentile explained that Jaspersoft already has about fifty OEM customers that use Jaspersoft BI in SaaS deployments, though to date, those efforts resulted from customized consulting work.

With version 3.5, however, those customization will be coming into the main product to enable a wider array of SaaS vendors to deploy BI alongside their offerings. For instance, Jaspersoft is part of Salesforce.com's AppExchange, which enables users to use BI with the SaaS service.

Though Jaspersoft is aiming to help enable SaaS vendors to integrate BI into their offerings, Jaspersoft itself is not delivering its software in a SaaS model.

"I don’t believe putting BI in the cloud and telling people they can throw their data up and analyze it is a good plan," Gentile said.

Instead, he said it's more valuable to tie a SaaS offering into a vertical solution.

"I don't think that there is a general-purpose BI market," Gentile said. "You will see us continue to partner with those that have functional or vertical orientation to their solution where they want Jaspersoft as their analytics alongside their own solution."

Open source BI vendor Pentaho is also seeing value in the cloud and SaaS-based deployments. Pentaho version 3, which was announced in March, included an improved Cloud Computing Edition. Lance Walter, vice president of marketing at Pentaho Corporation, told InternetNews.com that there has been solid interest in Pentaho's cloud version.

"We've gotten a number of inquiries, both from end-user organizations as well as existing SaaS partners who currently host their own datacenters but are considering outsourcing the datacenter piece to the cloud," Walter said. "One nice thing about our architecture and cloud computing is that it wasn't a huge investment for us to do this."