With all of the data analytics technology providers out there, how on earth are companies in that space supposed to differentiate themselves? One company is doing it by claiming to have come up with an entirely new way of approaching data analytics. Called “data algebra,” the approach is based on a “flash of insight”—that all data can be represented algebraically, which has enabled the creation of what the company says is the “first universal platform for data.”
The company is Encinitas, Calif.-based Algebraix Data, and it claims to have demonstrated that mathematics is the way to unify data management across different data structures. I recently engaged Algebraix CEO Charlie Silver in an email interview, and I asked him how the concept of data algebra has managed to elude us until now. He cited an array of factors:
Even though the computer industry was born from the mathematical work of Alan Turing—the brilliant logician/computer scientist famous for cracking Germany’s “unbreakable” Enigma code during World War II—the computer industry did not evolve mathematically. Instead, for a variety of reasons, it quickly became the domain of electronic and software engineers. Of course, many programs and software products today use math to process data (mainly numerical data) to excellent effect. And over the years there have been a few efforts to model the processing of data in a mathematical fashion, but they were unsuccessful. None of them resulted in the formulation of an algebra of data. Not even close. One major holdup, not surprisingly, was how hard this was to do. Teasing out a universal data algebra was not a cakewalk. If it had been, an algebra of data would have been propounded decades ago and would quickly have become one of the foundations of software. Now that’s about to happen.
I asked Silver what capabilities Algebraix will provide to its customers two years from now that it doesn’t provide to them today. He said the response is a technical one, so we need to bear with him:
The answer is: Autonomous conversion of queries into their most optimal, most efficient structure, and autonomous optimization of the execution of queries with our patented partial reuse capability. This combination of data algebra capabilities will be delivered as a “universal optimizer” that dramatically accelerates query execution and response. There’s a really important underlying concept here: Algebraix’s math-based technology caches the computation—reusing it instead of repeating it for future similar queries—rather than caching the data. The first implementation of this much-needed performance accelerator will be throughout the vast Hadoop ecosystem, where speed is a chronic issue. Our universal optimizer should cut Big Data computing costs across the board.
On Sept. 15, Algebraix announced that it was adopting a business model similar to that of Hadoop/Hortonworks, so I asked Silver to encapsulate the strategy and significance of that move. He did so this way:
Algebraix and the team of mathematicians and engineers that created the algebra of data recognized from the start that what they were developing was a really big deal, with huge implications for the world of data—which means for everything. Medicine, science, space, industry, travel, finance, physics, retail, defense, aeronautics—you name it, data drives it. But along the way, we also realized that data algebra is such a big deal that discovering and implementing all of its applications and benefits will take decades, and will require far more resources than Algebraix has today. That’s why we decided to open-source our mathematics. The goal is to encourage, accelerate and rapidly expand the countless potential applications of data algebra. At the same time, our young company and the private investors behind it have made a huge investment of time, money and talent—not to mention patience and persistence—in developing data algebra, and are more than ready to start saying these three letters: ROI. That’s why we’re adapting a version of the Hadoop/Hortonworks model by open-sourcing data algebra. By that I mean, as we transition from an R&D company to an active commercial enterprise, we’re open-sourcing our mathematics (a la Hadoop’s free programming framework), but we’re also patenting our own applications of data algebra, developing products based on them, and marketing our expertise in it (a la the highly successful Hortonworks Enterprise).
I mentioned to Silver that Shashi Upadhyay, CEO of Lattice Engines, said in my recent interview with him that the demand for data scientists is so high, and the supply so low, that “we are on the precipice of a data scientist recruitment war.” I asked Silver for his thoughts on that, and how Algebraix has been able to recruit and retain the data scientists it needs. He said there are two answers to that:
The first is that highly skilled people want to join cutting-edge technology companies. Algebraix is the epitome of that. The second is based on our experience alone so, yes, it’s a small sample but it represents something fascinating that we heard repeatedly while interviewing—and hiring—highly desirable data scientists from big corporate positions, with titles and salaries to match. Not only did we attract excellent candidates, but many of these bright, extremely well-educated professionals told us they were bored out of their gourds. It turns out that when they join major corporations as part of an in-house data scientist team, it doesn’t take long before they find themselves doing the same things over and over and over. So when we offered them jobs with the company’s key product, Algebraix Analytics as a Service (we outsource business analytics and include data scientist support), they basically jumped at the chance. Why? They get to work with an ever-expanding array of clients, each with different problems and perspectives, and needing different solutions. For example, right now our data scientist/analyst team is working with a top global educational group, a very successful California restaurant group, and a sports enterprise. Are they bored? Not a chance.
I asked Silver if he could have one do-over as CEO of Algebraix, what it would be. He said he would have focused more on smaller projects:
Not ones that take multiple years and require raising many millions. We started out to build a better database—a much better one—and we built an amazing one, but underscore “one.” And it took a ton of money and time. Since then, we’ve switched that part of our business from making a database to making databases better. Look for that next from Algebraix.
Silver wrapped up the interview with some advice: You should start turning yourself into a data scientist:
When I heard about the latest giant layoff at Hewlett-Packard, my first reaction was that DBAs and IT staffers everywhere need to reinvent themselves. Because part of what HP is doing is what major companies and tech divisions everywhere are doing: streamlining and automating everything they can. Rather than having huge staffs that take care of machines and tend databases, enterprises are investing in smart software that increasingly lets these huge systems monitor and maintain themselves. Sure, some human oversight will always be needed, but IT jobs are going to shrink dramatically. The only question is how fast.
A contributing writer on IT management and career topics with IT Business Edge since 2009, Don Tennant began his technology journalism career in 1990 in Hong Kong, where he served as editor of the Hong Kong edition of Computerworld. After returning to the U.S. in 2000, he became Editor in Chief of the U.S. edition of Computerworld, and later assumed the editorial directorship of Computerworld and InfoWorld. Don was presented with the 2007 Timothy White Award for Editorial Integrity by American Business Media, and he is a recipient of the Jesse H. Neal National Business Journalism Award for editorial excellence in news coverage. Follow him on Twitter @dontennant.