Hortonworks CEO Rob Bearden has a pretty good answer if you put it to him that Big Data needs to supply some more case studies and show there is more to it than just promise. Based on current developments, Bearden says, it’s possible that we can find a cure to cancer in eight to 12 years if medical researchers capitalise on the incredible ability to access and analyse data. (Hortonworks supports and participates in a range of initiatives designed to accelerate genomics and medical research, including Joe Biden’s Cancer Moonshot Task Force.)
It’s that sort of opportunity to answer the world’s most pressing questions that has so many of us cheering on this movement. But it’s surely true that we could use more evidential weight if Big Data is not to descend into Gartner’s dreaded trough of disillusionment that follows the peak of inflated expectations.
“Use cases are just taking off,” argues Bearden who speaks with a soft accent derived from his upbringing in the US south and has the polite formal manner characteristic of the region. “The correlation is one of the more data you put under management, the more you can drive value faster.”
The prospects for Big Data are growing and not retracting, he insists, as more customers work with high-volume analytics all the way along the value chain from the point of creating data rather than focusing exclusively on critical junctures later such as transactions. By tracking all events and processes throughout their life cycles and executing against pre-set rules there will be unprecedented access to information and analysis in close to real time, he argues.
“[Data analytics has] always been possible but not in true real time,” Bearden says. Capturing vast volumes of data from an array of sources then querying them on the fly will be enormously important in everything from beating diseases to transforming banking, countering fraud, automating the world - and at the same time possibly building the next software superpower.
“It’s the biggest transition at the data layer since the relational database emerged,” he says. “And it’s about being able to create a modern data architecture based on Hadoop that goes from the point of origination to data at rest and [the chance to put it] under analysis at any point with machine learning and very deep historical context.”
Hortonworks moved very swiftly to become a $100m annual revenue company on the back of its promise to translate the Hadoop open source framework to real business but its progress hasn’t all been smooth since. Bearden added sales to his responsibilities after then president Herb Cunitz left the company. Shares in the company have failed to rally even if this remains a sizeable concern and with plenty of capacity, Bearden insists, to execute on its remarkable ambitions.
He insists that the company has unique advantages over its old sparring partner Cloudera with a strong security governance architecture that includes work on Apache Metron, a security threat analysis platform. Hortonworks also enjoys a tight relationship with Microsoft.
“We’ve clearly passed them in terms of customer numbers,” Beraden says of its fellow Silicon Valley rival. “They’ve gone up the stack to compete with the core ecosystem and they’re aggressive with Amazon, IBM, Teradata… I see them diverging more to the pure analytics space and we’re more about the mission-critical data at motion and at rest.”
We’re much more cloud centric than they are, orders of magnitude, and we’ve got a much better ecosystem.
Bearden is also defensive when I ask whether in retrospect Hortonworks might have bided its time rather than jumping onto the stock market late in 2014.
“We were early to IPO in the traditional sense and it was about bringing validation to the technology and the space. [We expect the] entire data layer to re-platform around Hadoop and we see this as being as big a value creation exercise as ERP in the Nineties and we believed it was very important that the CIO [of a customer] saw the company they were dealing with was a public company. But you can never be defined by your stock price. It’s about your technology and what it enables and the transformations it enables. Will you take some short term hits? Yes [but] these are the early days.”
PREVIOUS ARTICLE«Mendix CTO on a friendlier IT department in transition
NEXT ARTICLETypical 24: Bart Simpson, Autodesk»