Pentaho is a traditional BI analytics vendor that has moved aggressively into the era of big data. It is trying to create a place for itself as the interconnect among transactional RDBMS databases, high-performance analytics databases, NoSQL databases and big data systems. If, as Ahbi Mehta says, the real value is in business analysis that works across data types and combines traditional rear-view BI with big data predictive analytics, then Pentaho has chosen the right course.

We provide a complete business analysis suite …. and have invested heavily in supporting every version of Hadoop as it came out,” Pentaho Technology Evangelist Ian Fyfe said in a live webcast from the Cube (see full video below) at last week’s O’Reilly Strata 2012 Big Data Conference. “And we go beyond that to support the NoSQL database space – Apache, Casandra, HBase, and so forth.”

“We recently put out a press release because we really believe we support more different data stores in general than anyone else. It is hard to find something we don’t support,” he told Wikibon Chief Analyst David Vellante and SiliconAngle CEO John Furrier. “We were the first analyst company to support the original version of [open source] Hadoop.”

Pentaho sees three markets for analytics today and sells into all three:

1. Traditional business analytics using relational data from a high-performance analysis database like Vertica.
2. >Big data predictive analytics drawing on unstructured data from Hadoop or NoSQL databases.
3. The OEM embedded market of software vendors who want to include analytics in a larger application – this represents 40% of Pentaho’s sales.

But the ultimate market it sees is for analysis that combines data from multiple database technologies. “We’re not just a silo. We’re not for instance just a Hadoop analytics vendor like some of our competitors. They focus on Hadoop and do a good job. But when you try to go beyond Hadoop you kind of stop. We want to provide this enterprise-wide view across big data and traditional data.”

Pentaho has also embraced open source, moving to a prescription model for its marketing that combines open source products with commercial products and services. It also supports both public and private cloud implementation and traditional on-premise, stand-alone server environments.

“We want to provide very useful software,” he says. “if we don’t, people won’t use it. A small percentage of users are looking for more functionality, capabilities, mission-critical support, services, etc. that’s where we make money, that commercial up-sell for people looking for additional capabilities.”

The market is evolving rapidly with the advent of big data, Fyfe says. A year ago most people at Strata were “kicking the tires”, trying to understand what Hadoop is and why their companies should invest in big data. This year those attendees understand why big data is important to their companies and are instead asking practical questions such as how do they go about integrating Hadoop with the rest of their infrastructures and how they can do analysis that combines big data with their transactional and historic RDBMS databases. And, he believes, Pentaho has an important and unique part of the answer to those questions.