For some time, companies have increasingly been looking to the Hadoop ecosystem as a way to scale their data warehouses and power business intelligence (BI). But while Hadoop can handle the scale, getting performance out of BI tools on Hadoop has been an issue.
"When they test Hadoop out and try to make that move, they're seeing that it's not performing at all like their Teradata or Netezza," says Dave Mariani, CEO and founder of AtScale, a startup that specializes in BI on Hadoop, with backing from Yahoo! co-Founder Jerry Yang and Cloudera co-Founder Amr Awadallah.
Business intelligence and analytics specialist Tableau Software is turning to AtScale to help it address the problem.
Enterprise customers have struggled with connecting Tableau to Hadoop, citing performance issues, security issues and scale issues, says AtScale Chief Marketing Officer Bruno Aziza. This, Mariani adds, has allowed big data analytics vendors to sell against Tableau with claims that Tableau doesn't work with Hadoop.
"That's a threat to those accounts, frankly," he says.
At its 7th annual customer and partner conference in Seattle in September, 2014, Tableau CEO Christian Chabot pledged the company would spend more in R&D in the coming two years than it had spent in the previous 10 years. Performance and elements like resiliency, high availability and security were among the key areas of focus outlined by Chabot. Cloud, mobile and data preparation also made the list, along with improvements to core visual analytics capabilities and storytelling.
AtScale, which focuses on providing business users with fast, secure self-service BI on Hadoop, announced the new partnership with Tableau on Monday at the Tableau Partner Summit. AtScale will deliver the industry's first "BI on Hadoop" accelerator with a special program for Tableau customers that Aziza and Mariani say will give business users the ability to run analytics on Hadoop at 100X the performance, with enterprise-grade security, without requiring IT departments to move, transform or sample data to make it available for analysis.
The key, Aziza says, is to power BI on Hadoop without forcing customers to change their BI layer or even move the data.
"This let's Tableau work on Hadoop without having to buy a verticalized, one-trick pony that only works on Hadoop," Mariani adds. "Enterprises don't want to have one analytics tool for Hadoop data and one analytics tool for everything else."
"Now they can land their data in one spot but still open access to that data in a secure, performant, controlled fashion with semantics that make sense to business users," he notes.
Under the partnership, AtScale will be part of a bundled offer to Tableau enterprise customers sold through Tableau's channel partners. Tableau's partner ecosystem will provide a discount on the bundled offer and deploy it within those accounts, Aziza says.
Sign up for Computerworld eNewsletters.