VERILYTICS IS A relatively small company trying to pioneer a big space. The company has developed a set of real-time data analysis tools that companies such as Avon, Household Finance, and Cisco are using to support their e-business initiatives. In an interview with InfoWorld Editor in Chief Michael Vizard, company CEO Shikhar Ghosh explains why it's a lot more important to figure out how you are going to use data than it is just to collect it.
InfoWorld: What exactly does Verilytics do?
Ghosh: We are creating a platform for solving the problem of taking information within a company and making it useful to individuals who use it. The space we're in is "e-Analytics." What we do is take lots of information, connect into back-end systems, and then apply a set of analytical modules to distill that information down so that people can then use that information. We have a way of taking everything the user actually uses, and the rules that they apply, to create a data warehouse on steroids at the back. It's all done in memory. We draw in only the information from other sources that's actually being used by users to do analytical work. Because of that, the scaling properties are much larger and we can have real-time information that gets used.
InfoWorld: Where is this tool most applicable?
Ghosh: The application in which we're starting to apply the technology a lot is in the financial services area, especially for people who need real-time aggregation analysis of lots of data to do their jobs. We're also seeing some applications in the consumer space where they have a lot of information that they're trying to get out to either their sales reps or to a broad consumer-base. We're seeing the use of this information in places where information is changing a lot, but people don't have the time to keep watching it -- places like a business-to-business exchange, where you have an end-user who wants to see what's going on there but doesn't have the time to manage it.
InfoWorld: What are the specific products?
Ghosh: There are three products that we have, and one that is still being built. The first one is a rendering engine that takes a variety of information a user might need, provides a set of personalization criteria, and then delivers that to the user. The second product is an alert engine that sits on the network. It looks at all the data sources that are coming in, and when certain events occur, it sends out an alert. The alert engine has behind it an analysis engine. What we're doing there is taking all of the information from the back-end sources and reclassifying them according to what users actually do. And then we have, for want of a better word, a language by which this information can be combined. The final product that we have is a text-processing engine. We have a prototype of it. The intent there is to combine structured and unstructured data and do a set of knowledge-discovery things on it.
InfoWorld: How do your customers make use of all this?
Ghosh: A quick example is how we're working with a very big investment fund managing 700 commercial portfolios using this technology. They have approximately 6,000 rules that they apply against all the data. The whole model that we're moving to with these products is a model in which there's more information than people can reasonably process. For example, Avon is using it in an application for their sales reps. Avon has 500,000 or so reps in the United States, and these reps require a whole range of information from the company. Avon wants to give them other information that is not necessarily Avon's, so that the reps come back and have more loyalty to Avon as a company -- because the reps are not employees of Avon. What Avon has done is create a personalized portal using the technology for all of its reps.
InfoWorld: How big an issue is managing data going to become?
Ghosh: What we're finding is that people are struggling with the problem of how to handle all of this information, how to make this information relevant and useful to other people. A great deal of what was done over the last 10 years has been solving the problem of accessing information. But once you've done that, the bottleneck becomes how somebody uses that information. That's the problem we're trying to solve.
InfoWorld: What's preventing that from happening?
Ghosh: One of the challenges we have is that most people who have looked at this before said the integration problem is too big. What they found was that the software takes six weeks to install and the integration takes another 12 months before they can make it useful. That's been one of the big barriers in getting past the notion of how expensive and how complex and how time-consuming the whole integration is going to be. These are some of the challenges that we're facing on a day-to-day basis.
InfoWorld: So what has to happen for people to really understand how to solve this problem?
Ghosh: The first thing that needs to happen is that companies need to start focusing on the problem of how much of their information is actually yielding business results -- how many decisions are being informed by the information infrastructure that they have built up. A lot of companies have an emphasis on creating the information infrastructure and very sophisticated tools to bring all that information in. The next big thing they've got to do is ask the very hard question of what they're really getting for this.
