A Talk by Eeshan Chatterjee, at Unconference
This talk delves into the nuances of building a data science team, the kind of people you need, whom do you start with and when do you start. We will go into the different kinds of data scientists and talk about which ones you need, and how many of each do you need.
There are very few companies that have truly been able to benefit from the vast area of data sciences. The major reason, is that they hire to solve a subset of present problems, but in the course, forget that a strong data science team can not only make a few problems go away, but also sometimes solve problems that might give the company a market niche. The best examples in the Indian startup ecosystem would be Housing, that shot to fame and captured >40% of the property search market share in an unbelievably short time. Flipkart, and it's strong focus on data sciences, has kept it's lead by launching new data products and tailored packages despite a lot of competition from big houses like amazon and softbank backed snapdeal. This talk delves into the kind of experiences and expertise a company needs to look for when trying create their data science teams. We will look at the different personality types and identify which ones serve you the best.
I'm an algorithms researcher and an data sciences consultant. I started out with Mu Sigma's Innovation group, working primarily on unstructured data mining before I moved to an internal consulting role, leading the Innovation Commercialization Team. I designed data driven & analytical solutions for Mu Sigma's fortune 500 clients. From Mu Sigma, I moved on to MediaIQ Digital, a digital adserving and analytics startup, where I try to bring together every available dataset that can describe an individual better, and build intelligence around targeting users with non-invasive, contextual advertisements in the real-time bidding framework. I also help many startups with problems in the data sciences field, in particular bootstrapping their data efforts and teams. Check out my linkedin and github here: