Hadoop came out six years ago enterprise-ready | #HadoopSummit

pink_elephant_in_glassesBeing the first company to architect their entire suite of software solutions on the Hadoop platform, Tresata is, it is safe to say, fully committed to the technology. Abhi Mehta, the CEO of Tresata, has witnessed the #HadoopSummit grow in popularity and attendance over the many years he has been coming. And he sees that steady drumbeat for the technology as a very good sign.

“It is refreshing, in a positive light, that the journey we started together in forecasting the future of data, not the future of technology, but the future of data and how it empowers enterprises to re-do business models is well underway,” Mehta noted. Continuing he said, “We are finding practitioners finding incredibly relevant and transformative business value by taking data, not just from one source or two sources, but multiple sources, combining it into data assets and solving problems that literally could not be solved before.”

Expanding Applications On The Hadoop Platform

Even though he sees Tresata as being an instrumental partner for many of the early adopters, he concedes the market is still lacking in terms of application development for the platform. Even still, he recognizes an important aspect of this year’s summit he views as particularly distressing. “I’m upset that data transformation is the biggest part of this conference,” he stated. “Data transformation is free. We give data cleaning, data curation, data parsing software as part of our application bundle away for free.” He believes it to be the wrong conversation to have altogether. “You cannot expect a business user to sit around and have a user friendly tool to clean data. They expect clean data.”

Watch the entire theCUBE interview here:

By offering the full range of tools to their clients to operate on the Hadoop platform effectively, Mehta believes the paradigm for software sales is in mid-shift from the predominant model of only just a few years ago. Gone are the days, he explains where a backslap and a pair of tickets to the next golf tourney will ensure a sale. The entire operation needs to know that the sales professional has not only an understanding, but a commitment to delivering the best outcome possible as a result of adoption. Mehta also sees dark clouds ahead for tool companies whose aim and focus are to address one issue on the platform.

“We are building Tresata for the long-term,” he explained. “What are other people priming their companies for? Why people don’t understand that tools and infrastructure will get commoditized baffles me. But I think if you are not building a company around predictive analytics on a stack, it will not work. If you are building a tools company, it will not work.”

He backs this up by talking about how each tool is critically important but new companies will not be able to build value through offering them. “Those will be components of larger products that are offered for free. It doesn’t work when you offer it piecemeal through multiple vendors.” He concludes his doomed take on this subset of the industry saying, “ETL is free. BI is free. Data transformation is free. Metadata is free. Those are functions that you can’t build customers around. Give it up.”

Hadoop Is Not A Storage Platform

Echoing statements made at last year’s Hadoop Summit, Mehta cautioned against thinking of Hadoop as a storage platform. “It was built to be a massively parallel computational engine,” he explained. The next chapter, and the entire raison d’être, for big data and application development geared to it will be to eliminate manual business processes. Mehta discussed a task that has typically been a monumental human challenge: finding money laundering. Prior to the advent of the Hadoop platform and specifically engineered applications, this process required as many as 4,000 individuals slaving through tickets trying to determine which transactions were fraudulent and which were not. “What we are doing now is taking it to the next level,” he stated. “We are automating processes in thing like deep learning, artificial intelligence, algorithms and machine learning and taking business processes that required thousands of people and bringing it down to tens of people.”

Mehta recognizes there is still trepidation on the part of enterprises pulling the trigger and committing to the Hadoop platform though he doesn’t fully understand why. “Hadoop came out six years ago enterprise ready,” he said. “It’s always been enterprise ready. Let’s not forget that there are companies, incredibly profitable next-generation companies who have architected new business models on Hadoop.” Referring to companies like Google, Facebook, Yahoo, Microsoft, Twitter and LinkedIn, Mehta stated, “So, when someone says Hadoop isn’t enterprise ready, are they telling me that Google, with a $300 billion market cap, isn’t an enterprise?”

Mehta predicts the convergence of cloud, big data and mobile will eventually help to ease the pain of transition many in the enterprise set are anticipating they will feel. Ultimately, the convergence will make it plain that not only will the transition make the performance of their day-to-day business operations easier, but that it will be an accepted eventuality that simply has to happen.

Where the greenfield operators like the Google, LinkedIn and Facebook came in and created their business models around the new technology, Mehta acknowledges the challenge faced by legacy and offers his two-cents on how that challenge can be overcome. large_12119363“I think the biggest challenge for legacy is finding a nugget and a team and someone with balls to take some risks, bold moves that give you early wins,” he stated. “That is the biggest issue for legacy not being able to deal with the disruption.” If someone within the organization is daring enough to follow Mehta’s advice, he hopes the company will recognize the value of the risk and support them to success.

The Hadoop field has seen supporters jump in the water representing new start-ups, established start-ups and legacy vendors. With legacy vendors being termed the whales of this year’s summit, Mehta offered his prediction for who wins and loses as the market continues to mature. “Give me a pack of piranhas versus a whale,” he said excitedly. “I will pick piranhas over whales because when there is blood in the water it only ends one way.” He concluded, “[Hadoop] is a replacement for the technology stack. It is not a complement.”

photo credit: pink elephants pics on Sodahead
photo credit: willc2 via photopin cc