Sramana Mitra: Is there anything else you would like to comment on in that domain?
K. R. Sanjiv: As far as B2B companies are concerned, I think it is not so much of a fantasy. The potential they have today in terms of white spaces within their business processes is phenomenal. There are a few things I see and that customers have realized. One is that big data is to be looked at from a business-backward perspective rather than a technology-forward perspective. We have seen a lot of customers that ended up spending a lot of money and time, etc. Then it comes out and people say, “This is it,” rather than looking at it from the other perspective by asking, “What are the top five business problems?” The second thing we have seen is that a lot of people are trying to use new technologies that were designed for a purpose. It requires a lot of engineering to make them work across enterprises. There is a lot of work involved, and people often don’t realize that. This is another point in big data where we see a lot of customers fall into the trap.
SM: As far as your workforce is concerned you said you have about 8,000 people in the organization. One thing I am hearing from people in your field is that there is a shortage of good data scientists. A data scientist should have a unique combination of math, statistics, technical understanding, and business acumen. That is a combination people have a real difficulty finding. How much of that talent is a requirement in the way you go to market and deliver your solutions? Is that a problem you are facing.?
KS: Absolutely. The way to address this problem is to put the three in a box. If there is a problem to be solved, we have one mathematician, one domain expert and a technology expert together as a team. The need is for people with a scientific and statistical background. Data is meaningless unless you have some knowledge of the data, and someone who knows the business processes around the data and who can sense and detect interferences with it. That is a breed which is heavily in demand. We are trying to train these people internally, but at this point the approach is to put three different people in a box with three different dimensions and try to address the problem.
SM: What you are saying is that the composite of those is very difficult to find.
KS: It is very difficult. Historically, the way analytics was done is that there were five people sitting in a group inside a department doing the design of experiments. Now it is much broader. You need people who can work across the enterprise. It will take four to five years for that gap to be filled.
SM: Where is most of your big data team?
KS: We are a team of 8,000 people. About 35% of them are globally distributed and 65% of them are in India. In terms of statistical knowledge, we have put a lab in Calcutta, because that is the city where most of the statisticians are available in the country. Now we are running out of capacity. So we are looking at countries like Romania or China, which also have a pool of statistically trained people.
SM: How big is your Calcutta statistics team?
KS: Around 250 people.