Jim Stikeleather: So, that’s the area where I think there’s tremendous opportunity. How do you embed real-time analytics into the data stream rather than using the after-the fact-analytics most people use today. Then, how to reintroduce humans into this analytic process so that they have some oversight over decision making. My suspicion is that the form this is going to take will be multi-sensory output devices, devices through which sound, feeling, and visual and three-dimensional [all work together]. It’s kind of like a Nintendo 3D interface replaces the iPad tablet interface if you will, so that humans can take advantage what we do best, which is use multiple senses to identify patterns that we don’t generally identify terribly well just using one resource, for example, our eyes, which is how we do things today.
Sramana Mitra: Embedded real-time analytics as a complex architecture is a complex business model, one with many levels. Moving from describing the term to turning it into a business model; it’s complex at many levels, so let’s go through some industries where there is at least some beginning of embedded real-time analytics. I can think of advertising, targeting advertising, as one of the areas where there are probably the most immediate applications. Is that an accurate observation?
JS: I think target is real-time advertising, and if you think about the model we seem to be moving toward, consumer computing where consumers are going to be focusing on phones and tablets and probably less so laptops and desktops. Well, almost all of those devices maintain some sense of GUI presence so much of consumer computing is gone to an advertising support model. Google can give you free email, and most of the apps available on Android and iPhones are basically free, so now what you want to be able to do is in real time identify who the person is, where they are located, what’s around them, and who they are with. In effect its this massive data-crunching effort. We want to look back to past behavior to be able to say, Hey I see you like Chinese food. You just went to the ATM machine; here is the coupon for the Chinese restaurant just across the street from where you are walking right now. That is the next step beyond just a Google inquiry banner page popping up. So that’s one of the places we can go.
SM: What have you seen so far in terms of interesting entrepreneur ventures? Do you know of innovative companies where there is good evidence of embedded real-time analytics?
JS: That’s a good question. I haven’t specifically been looking for those because the Dell focus is really on the systems and the underlying infrastructure. So, no, I can’t tell you that I’ve actually looked at. We have paid more attention to things like embedded real-time analytics inside systems management software. This would include looking at things like the law in real time. We see what’s happening in the environment and then make the appropriate adjustments to the system configurations – you know, setting up virtual servers, dropping virtual servers. The other place that we are focusing on in terms of this type of stuff is obviously in the security space. Looking at attacks as they are happening, looking at email flows. If you detect that this server here is suddenly sending the same email to a dozen different places, you probably should be blocking the email. So, that’s where we’ve been looking at the analytics. We haven’t been necessarily looking at the consumer space, although that’s probably the space we are going to see the most growth and the highest value applied to it.
SM: Yes, I think the spaces you discussed – infrastructure management, IT structure management, security – those are architecturally somewhat simple to manage because the enterprise still has control over the infrastructure and can make decisions on what goes on a particular device on the client side, how much of that crunching of real-time analytics happen on the device. I think on the consumer side, the problem is much more complicated with privacy and all that.