Platfora is continuing to evolve its Hadoop-focused analytics platform with a series of features catering to various user roles, such as data scientists and business analysts.
The company’s software stack runs on top of Hadoop, taking advantage of its processing power and scalability while keeping its complexity under the covers and giving users visual ways to interact with data.
Platfora uses the term “lenses” to refer to data sets that have been put together with its tools. Version 3.5, announced Wednesday, adds programmatic query access, meaning users can use programming skills such as SQL to push lenses into other systems, such as R for statistical analysis, or a recommendation engine, the company said in a blog post.
Business analysts who may lack such programming skills now have the ability to export Platfora lenses and other objects into a .CSV file.
Platfora is also rolling out a new plugin framework with the update, which customers can use to connect public sources such as the U.S. Census with private ones like enterprise applications.
Other improvements in 3.5 include more than two dozen updates to Platfora’s visualization layer that are aimed at making the user experience easier and more intuitive, according to the blog.
In addition, Platfora 3.5 introduces updated support for Hadoop distributions, including Cloudera CDH 5 and MapR 3.0. Hortonworks HDP 2.1 support will arrive next month.
Platfora is fighting for attention among many other vendors looking to exploit the explosive popularity of Hadoop.
At this point, the industry has laid out a basic infrastructure for Hadoop, said Platfora CEO Ben Werther in an interview. “Now we’re focused on how do you unlock [its potential] for the average business user?”
“The conversations we’re having now are almost always line-of-business focused,” he added. “That’s the big shift now. Companies going around trying to sell an IT solution around Hadoop are in the wrong conversation.”
Platfora 3.5 will be available Friday and is sold via subscription on a per-server, per-year basis. Specific pricing wasn’t disclosed.
Companies like Platfora tie into a growing desire among customers for “agile” analytics that help users find answers to questions more quickly, said Forrester Research vice president and principal analyst Boris Evelson.
With smaller data sets, this type of agility is possible with desktop-based in-memory applications such as Tibco Spotfire, he said.
However, “when you have terabytes and petabytes of data, that’s where Hadoop and tools like Platfora come in,” Evelson added.
There are some compromises with the agile approach, such as the need for extensive data cleansing to take a back seat, Evelson added. For tasks such as regulatory task reporting, which require great precision, traditional approaches to data preparation must be taken, he said.
But for jobs such as customer sentiment or customer segmentation analysis, “it’s good to have clean data but what’s really important is to have speed to market,” and that’s where Hadoop and tools such as Platfora come in, Evelson said.
Platfora’s analytic stack is “interesting and innovative,” said analyst Curt Monash of Monash Research. Beyond its data-visualization and discovery capabilities in conventionally relational and multidimensional use cases, the platform can be used for analyzing event series, for applications such as online marketing, online security, and anti-fraud, he added.
Event series are sequences of records, often from several different data streams, that can best be understood when analyzed together in time order, Monash said.