By Sramana Mitra and guest authors Shaloo Shalini and Bhavana Sharma
DF: The other area I think of as white space is, if you think about the cloud, if everything plugs into the cloud, it’s like an economy. You need people to be able to make that market, and there are lots of impediments to doing so. People describe their services differently, people request them differently. I think there needs to be something that’s almost like Froogle, the Google Froogle, but for cloud services. A tool that has gone out and looked at these services, indexed them, and tagged them using the business service vocabulary and aliases – basically something like knowledge management or search, like we are searching for APIs about pages.
If you look at the way a tool such as Google works, it’s gotten to the point where when you ask for something, the search engine knows all the synonyms for it. If you ask for a telescope, it realizes that you might have actually been looking for binoculars. You need to be able to index and alias, and something is missing in cloud computing for this. The term we use in rule systems is “business vocabulary,” where you can express things in nontechnical terms. Moreover, you need brokering because a lot of this stuff has happened. Say, for example, I want to spin up a test or I have kept certain requirements where I don’t want to pop up a window to the person, and say I did the Google search and I want to tell that person, here is the set of results; do they look OK? I want to do this search under program control, pick one, and do it. This funny mix of search and binding but with direct punch-through – you basically wrap the request – that is a space where nobody is yet. Also, there is an economic aspect to this. Today cloud service prices tend to be relatively static; they’re not driven by client demand, and so I think something that understands how to do the arbitrage or make economic decisions comprises another white space.
SM: You are effectively looking for a vertical search engine for these programmable APIs, right?
DF: Yes, but it is certainly more than the search engine. It may be the vertical engine. Another way to think about it is maybe it’s like Expedia for cloud services. It is the nature of the white space – if I could describe exactly what I wanted, then it would not be a white space.
SM: Yes, somebody needs to visualize how this is going to pan out.
DF: Right. Basically, the white space tends to be three dimensional. When you are in the white space, you are lost in the cloud already. It surrounds you like fog, and you have no idea what you want to do, where you are going, or where you are. People ask me all the time, When I am in the car, where am I? You know where you are in that you can say, Yes I am in a car on planet Earth. It is the details in between that are always fuzzy. It is similar to being lost in a fog. I don’t really know if it’s something out there. I am a big believer in the concept of convergent evolution. In the evolution space you see this. If you take a look, you may think that an otter and a platypus may be related because they both have fur, they live in the water, they eat crayfish, and that kind of thing. But they are radically different because they came up in different domains. I am this believer that in the computing domain we tend to have evolutionary pressures so to look in the way economies evolve the physical economies – the way the retail industry evolved, about you get aggregators, you get the people who classify and normalize things to some extent what a book store Amazon does. And then there is an economic model, there is a business model out there for it. I just don’t know exactly what it is.
SM: And that’s what you are looking for essentially?
DF: Well, for white space, yes. You are asking me about white spaces; there are white spaces that I see in general, but that doesn’t necessarily mean those are the types of white spaces we want to move into. We do certain things in IT. The second part of the question is where we think we need to invest ,whether it’s to do acquisitions or development. I would say if you look at our products, one of them is in optimization. One of the terms used in IT management is “informed automation,” which is when you monitor and analyze information, and based on what you find you make an automated change. So, you are monitoring an environment and you detect that your online self-service application is running out of capacity. You want to automatically provision more machines, you do not want a person to do it – this is called informed automation. I use the term informed optimization because what you really want to do isn’t to make an automated decision; you want to automate the right decision. If you are automatically performing the wrong task, what you are doing is screwing up rapidly and in a repeatable way. An area we are investing in is optimization, which is look at the environment and make optimal decisions. That I think is a place in our product portfolio where there is a white space. But the nature of the white space, especially in enterprise software, is you want to get there, but you don’t want to get there too far ahead of your customers and in advance of when they would need such a solution. We want to stay a few steps ahead of what customers need, but we don’t want to be twenty steps ahead.