Jim Stikeleather: So, the data are “talking” I have transactional integrity out of the communication trail, and all those kinds of things. I think this will be offered as a service because over time, unless something has a specific need to be running in a dedicated environment, I think you are going to see almost all organizations running hybrid environments where they have a mix of traditional types of computing applications, maybe email and collaboration applications, maybe NFI communications appliances, private clouds and public clouds all in a mix. Somebody is then going to perform and manage the integration among all that.
Sramana Mitra: I guess what I was referring to when I said services rather than products, I consider SaaS tool to be a product business as opposed to a service as business for you where you actually have to perform customs software work and stuff like that, so when I just as a point of clarification I wasn’t really
JS: I see.
SM: Thinking about this from an entrepreneurial opportunity point of view, you are saying this opportunity is a combination of software as a service and actual hands-off service?
JS: Software construction services, yes, because what you are seeing is that standards are coming out relative integration and communication, BPEL and BPMN. We have activities that are going on in terms of defining common business architectures, which are also defining common interfaces. So, there is a lot that will be in common, but there will always be a certain amount of tailoring. Probably SaaS plus a small adapter type of construction for whatever is specific to what the customer needs integrated.
SM: OK, got it. So, flipping topics, I’d like to discuss in more detail your point on big data analytics. Analytics as an industry has been around for a while, and there are major players in the business. I know from my experience and from this series we’ve had the CIO of IBM and many other people who are deeply involved in cloud services. IBM is going after that space in a very big way. Would you tell me more about your impressions on where are the gaps and where do you see that industry going? Where are the entrepreneurial opportunities?
JS: Well, I will describe what I consider the future scenarios and their implications and where I think analytics are going to end up. I think you have to keep in mind a couple of things. Let’s take the work mobility theme I was describing where for one transaction – I bring on a new employee – it goes to a dozen different companies to be executed on. So, now I have a new analytic space because the data that I want to analyze is spread across multiple companies. That’s a new dimension to traditional business analytics.
The other thing that is going on – I think everybody’s heard the term – if you think about the Web today, most of the content on the Web is still human generated, but as we start embedding things into our phones, into our refrigerator, our stoves, our cars, biomedical devices that are constantly reporting on you, even individual you know cans of food with RFID chips on in them, with all that we’re moving to an environment where an ever increasing amount of data is being generated. We are approaching what I refer to as information overload, and the point where the value of the data diminishes rapidly over time. You have to be able to know what you know, and the value of the data becomes time sensitive. You need this ability to not only get access to data across multiple organizational boundaries, but to make sense of it and use it because there’s much, much, much more data than we’ve ever dealt with before.
Because data has a certain amount of time sensitivity attached to it, you want to have embedded real-time analytics so that you have taken analytics and put them into the data stream rather than looking at the data after the fact, because the data stream is where you will see the actual storage of the data. This may be on another company’s machines because of privacy laws or something else. The data itself cannoy be stored on your machine. There are laws that are being passed like in Luxembourg on financial transactions and in Germany on medical transactions where you are not allowed to keep the data. It can flow to your transactional system, but you can’t store it, so you start moving into mass [use of] real-time embedded analytics. And they take advantage of it; you have to make the analytics have intentionality and make some form of decision; maybe it is the reorder points, maybe it’s price changes. But the problem there is you then have the opportunity for this to go very wrong. I don’t know if you remember in February the flash crash on Wall Street, but that was a classic example of embedded real-time analytics making actionable decisions without human intervention. We had a problem until a human stepped in.