Sramana Mitra: Absolutely, that makes sense. Who else in the real-time data access and analytics space plays in the ecosystem?
Ari Zilka: In the real-time analytics space, there is a lack of clarity as to what tools to use in the market. The vendors have started to paint fences and mark out territories like complex event processing and real-time business intelligence engines.
SM: TIBCO plays in that space, right?
AZ: Exactly. TIBCO plays in that space. Pegasystems plays in that space. So does Software AG, our company. Oracle, IBM, Progress Software, Red Hat, Flash, and JBoss – those are the main players at the moment. Essentially, there was a realization around 2006 that batch is not going to work. The reason it’s not going to work is because in order to answer some simple questions, like . . . Our customers ask, so we have one of the world’s largest credit card processing networks that runs on top of Terracotta. When you swipe your credit card, a Terracotta-based system flies into action. That system does looks at each of your transactions. I’m not saying across all your credit cards, but let’s say you have a Bank of America XYZ credit card or a Chase XYZ credit card. There’s a brand of credit card and then there’s a bank behind it. So, for your Bank of America XYZ credit card – I don’t want to say the name because I shouldn’t – but for your Bank of America credit card, what happens when you swipe that credit card is that the system looks for every transaction you’ve ever executed on that particular credit card and says, “Is this out of pattern? Is this the dollar amount you spend? Is this the store you shop in? Is this the city you shop in?” What products are you buying?” It looks at every aspect. Is it the time of day you go shopping for this kind of thing? It looks at every single aspect of your buying history and decides whether it’s going to accept this transaction or whether the credit card vendor is going to call Bank of America and say, “I think this credit card is a stolen card.”
Before Terracotta, this wasn’t real time, it was batch based. It took x minutes. I don’t want to give exact details. But it took tens of minutes for an Oracle-based database to crunch all of this data about your usage history and return the response, fraud or not. And then Bank of America would be sent a message: Call up this person and confirm if he used the credit card. If not, shut it down. We’ve all experienced those calls, right? Today, that call is given to you courtesy of Terracotta. Every transaction in your use of this credit card network is stored in Terracotta, and the tens of minutes have turned into four seconds. They are now capable of detecting fraud before the fraudster leaves the store, which is very powerful.
SM: Interesting, very interesting.
AZ: That’s all about moving from batch to real time. There are two ways to do it. There are package products from TIBCO, Oracle, IBM. And then there are hand-built solutions on things like Terracotta. We also have a packaged product. Why are people building by hand? They’re confused. They don’t know what the shape of the new data management layers should be and how cloud and data management affect each other. But why are they building it? It’s because they want to move to real time.
This segment is part 3 in the series : Thought Leaders in Cloud Computing: Ari Zilka, CTO of Terracotta
1 2 3 4 5 6 7