Think Big Data seems to be the running theme at IBM’s IOD event this week, and Wikibon founder Dave Vellante, along with SiliconAngle founder John Furrier were able to speak with Tom Deutch, Program Director for Big Data at IBM about his thoughts on big data and what direction IBM is moving toward during Big Data Week. Vellante emphasizes the conference theme of “Think Big” because there is whole sea of information out there about big data.

It wasn’t too long ago IBM transitioned into services and software, with their main focus on software, such as Lotus. The company decided to focus more on software so that it had the ability to drive value to the customer.

One of the ideas Deutch spoke a lot about was a “Fit for Purpose” paradigm, which focuses on how to condense data to make it all fit. There is a NoSQL movement that allows us to consider developing a “fit for purpose” method.  This approach matches the compute problem to the best underlying compute technology, which results in a break through for analytics. Business outcomes are driving this, and the idea is to match the correct technology for the problem at hand.

Over the past 18 months, Deutch has seen a transition from exploratory systems to “how does this fit in with the enterprise?” which, showcases the evolution of technology.  Nowadays, businesses are looking at how they deploy IT, which is what Hadoop is doing.

The database world was boring at one point in time, but it’s now on fire. We are now seeing a trend in movement to specialism and speciality databases.

Deutch explained how if real-time analytics and at-rest systems, like Hadoop, combine, then that creates an adaptive and dynamic flow of information. The whole idea is to not pick from real-time or at-rest systems but rather use both of them together.

IBM has been quite successful financially. The company even beat out Apple in the stock market, which is no surprise since it had the biggest monopoly the computer business had ever seen.

To see Deutch’s interview in its entirety, see the full video below: