Sramana Mitra: The question I am asking is: What tools and technologies are you using to address that problem?
Venkat Viswathan: The tools and technology case is interesting, and again it is another area that has been going through a lot of change during the past 10 to 15 years. The biggest player in the space is SAS, which is the largest private software company. They continue to have a very strong market presence in this space. A lot of our large enterprise customers are well integrated with them. We tend to use that platform. But interestingly, the new kid on the block is a programing language that was developed 24 years ago in New Zealand, called R. It was developed by a couple of researchers who were far from mainstream technology innovation. It has been around 24 years, but it has really come into its own over the last 10 years. Much of this is driven by how it was adopted by PhD students at Stanford, Berkeley, or other centers of innovation in the U.S. As these students started working at Facebook, Google, or LinkedIn, they gravitated back to their comfort zone, which was using R.
R is an open source programing language, and it has created a significant momentum in terms of open source innovation of tools and technologies using the R dataset. Historically, it has had technological limitations in terms of in-memory processing and in terms of how it is able to handle large datasets. But in the past five years, during the emergence of big data, more and more innovation has gone into that, and a lot of the limitations are going down. Around a third of our work uses R as a programing language, and the tendency is rising. Then there are innovations around Python and the platform itself, which is allowing us to use open source platforms – wherever clients are comfortable with it – to do a lot of the unstructured data analysis we were talking about.
SM: How do you charge for these projects?
VV: Our business model is what I call an analytics center of excellence. What that means is we set up a dedicated team for our client. If you are Microsoft and you want to work with us for the next two years, you will work with a team of trained people. We assemble a team of 20 people with the right combination of business data and math, and the right mix of people onshore, embedded with the client, and people offshore in our global delivery center in Chennai. Then we draft the fine print of a contract, which allows us to run this team in a dedicated manner for Microsoft. What clients get is a set of talented people – a core team they can work with – and the ability to flex up or flex down based on business needs during certain quarters of the year. It is a scalable model in addition to the internal investments they have. They can essentially create a more scalable model in terms of the speed with which we are able to find people, train, them and then deploy them.
SM: They pay you according to the number of people you assign to them?
VV: Yes, it is a retainership model.
SM: The flip side of this equation is of course that you need trained, competent people in the customer mapping business. Where do you hire from, and how do you manage this kind of operation? How many people do you have?
VV: Just as we have to define what our market is, who the clients are that we serve, and what the problems are that they face, there is a supply chain problem to be focused on – in India as well – to identify where talent comes from, which markets are underserved, what kinds of skill sets we need, and what groups tend to have these skill sets better developed. That is the approach we take to find the right combination of business technology and math.
Today we are about a 200-person organization, but we are growing over 100 percent each year. There are only a few companies in India that have been there every single year for the last five years. The way we have gone about identifying, tapping into, and managing these talent pools is to apply some of the analytics we were talking about for our clients to some of our human capital operations. We do a lot of measurement in terms of testing. We have an online platform on which we do a lot of this testing. We also rank universities based on data we have collected about the curricula. Then we spend a lot of time in the field, meeting with students and visiting these universities. But having all this, in our current scale today, we only need three to four universities with which we build deep relationships for us to be able to meet demand.
We find there is the perception that India a single country. In my view, it is 25 countries melded together. There are vast differences in culture, in the focus on education, in consumption patterns, or in [wealth] aspirations. It is quite different in different cities. Our focus on Chennai is fairly deliberate. This I learned from our partners in Copyworld. What has given our partners a stable, scalable workforce is they picked a city where they have recruiters for this particular space. I see the logic;they have a very strong emphasis on math and are very good in U.S. cities, but [there is] not enough business hiring here. So, we are one of the top recruiters here.