Sep 22, 2013
Published By : Live Mint
Reams of data are generated everyday; these practitioners mine the information for cues on what sells and why
Analysing billions of data points and making sense of them is what data analysts excel at. Whether it is Web-browsing data trails, social media communication or even video or audio data files, managing these petabytes (10 to the power of 15 bytes) is what these data scientists enjoy doing most.
This data deluge (popularly referred to as "big data") is the new catchword as far as superior decision making is concerned, and experts with the computer skills to store this data, retrieve it by writing the right software algorithms and with the business analytics to use it, are in huge demand. We ask these professionals what it feels like to be in what the Harvard Business Review describes as the "sexiest job of the 21st century" and what it takes to keep up with the technology.
Srikanth Karnakota explains the intricacies of cloud computing with ease, making him sound quite like the evangelist for big data. "Organizations today have a big 'data' problem, they are able to make sense of just 20% of their data, the balance 80% is just there, and that is why big data today is one of the top priorities for the CEO (chief executive officer)," says Karnakota, who spends his days marketing Microsoft's Server and Cloud Azure, both tools for big data management.
How he got here:An electronics and communications engineer from the Jawaharlal Nehru Technological University, Hyderabad (1997), and a master's in business administration (MBA) from the Indian Institute of Foreign Trade, Delhi (1999), Karnakota joined Maruti Udyog Ltd, where he worked on exports and imports at the Nhava Sheva port in Mumbai.
"But the itch to join the digital side of the business was always there," says Karnakota, who applied in 2000 to Pramati Technologies, a start-up that made application servers. In 2004, Karnakota, who was handling business development for Pramati, joined Microsoft. He built and incubated start-ups for Microsoft and in 2008 moved to the cloud and server side of the business.
Daily duty:Karnakota's job is to sell and service the Microsoft range of tools and solutions for big data. These include the Hadoop-based Windows servers (the Hadoop software framework supports data-intensive applications) as well as Windows Azure. On a typical day in Hyderabad, he may sit down to work with his four product managers, to work out a sales incentive structure for Microsoft salesmen. Microsoft in India sells through a complex web of approximately 800-strong sales force.
Before we met, Karnakota had to speak at a big data event, organized by Microsoft, where experts had been invited to speak about the challenges and solutions of big data. Here, Karnakota says, he liaised with some clients and potential customers.
At office, there are a few internal meetings, which Karnakota conducts from his cabin via Lync/Skype. But virtual meetings notwithstanding, Karnakota travels a lot, both for internal meetings and meetings with customers. He spends three days a week on the road, mostly in Delhi, Bangalore and Chennai, with a quarterly visit to the Microsoft headquarters in Redmond, US.
Most interesting project: He worked with a car company that uses big data on consumer preferences derived from the Internet and social media. This was integrated with their dealer software. "This is where the magic happens. The dealer from Coimbatore now knows he has to store fewer jazzy colours because Coimbatore consumers prefer sober colours. The dealer from Nagpur knows his biggest customers are likely to be from the farming community, so he should design his promotion schemes to appeal to them," says Karnakota.
Biggest challenge: "Anticipating the trends that are going to shape the industry and being ready for that," he says. Karnakota keeps up with the latest technology trends almost obsessively. "You will be dead if you are not curious," he says.
Skill set needed for this field:"At Microsoft, we are a fairly opinionated team, so we look for people who are confident and believe in themselves. Also, people who are driven; else it is easy to get sapped out by the complexities of this business," says Karnakota. Decent programming skills, passion, curiosity, as well as the ability to collaborate and communicate are some of the qualities data analysts need to succeed.
Money matters: The salary at this level can be upwards of Rs.1 crore per annum.
Vinita Sivaramakrishnan has always been a fan of numbers. "I like data; I like making sense of what is happening with the aid of numbers," she says.
How she got here: After a bachelor's degree in statistics from St Xavier's College, Mumbai, and master's in political science and operations research from the London School of Economics and Political Science in the UK, Sivaramakrishnan joined Genpact's analytics division in Bangalore in 2010. In 2012, Sivaramakrishnan left Genpact to join Chennai-based consulting firm, Takshashila Consulting.
"That job (Takshashila) wasn't very data driven; I did it for a very short span of time before realizing consulting wasn't for me," says Sivaramakrishnan, who joined the Mumbai-based Hansa Cequity earlier this year. "Since the company is only six years old, it has a start-up feel to it. It also gives me the opportunity to work in retail marketing and to work face to face with clients in the Indian market," she says.
Daily duty: Sivaramakrishnan begins her day with a data study. "There are always business queries that require a deep dive into data," she says. The morning is spent trying to examine what caused a spike in last months' sales for a client. She also needs to track response to the clients' advertising campaigns and/or mailers. This involves working on SQL servers as well as Excel. Sivaramakrishnan has learnt coding on the job and sometimes writes her own code.
"C and Java used to drive me crazy when I was in school. But now when I know what output I need, like the figures on the sales of a particular month, it's easy for me to write a query and pull that information out of the data warehouse," she explains.
Most interesting project: Designing a strategy, while at Genpact, to optimize debt collection from credit card customers. "It got our client amazing returns and is being used even today," says Sivaramakrishnan.
Biggest challenge: "Learning to take that extra step to translate your statistical analysis into layman's terms to be able to communicate with marketing guys." Skill set needed for this field: "Quantitative skills; you can't flip when you see a spreadsheet with 1 million rows of numbers," she says. Money matters:The starting salary ranges from Rs.4.5 lakh-9 lakh per annum. With three to five years' experience, it can go up to Rs.7 lakh-15 lakh per annum.
"We are like Sherlock Holmes; we do what the human mind has always wanted to do—find answers to questions," says Srikanth Velamakanni, who quit a career in finance 13 years ago to become a co-founder at Fractal Analytics Inc., a data analytics firm.
How he got here:After a degree in electrical engineering from the Indian Institute of Technology, Delhi, and an MBA from the Indian Institute of Management, Ahmedabad (IIM-A), Velamakanni joined ANZ Bank in 1998, moving a year later to ICICI.
At ICICI, Velamakanni worked on designing India's first collateralized debt offering (CDO). "We had to do a lot of very interesting math behind that in understanding the risk profile of the different cash flows and pooling them in a way investors found attractive," says Velamakanni. The math was interesting but didn't seem to add much value to the world. So, in 2000, Velamakanni along with an IIM-A batchmate, Pranay Agrawal, set up Fractal Analytics "to deeply connect with consumers, earn customer loyalty, make better decisions to reduce waste, and ultimately improve lives".
Daily duty: Velamakanni's day in Mumbai begins at 8.30am, with Fractal's executive committee global telephone call—a seven-city call, where heads of the different offices, the chief financial officer, and the heads of human resources and marketing get together for an hour and a half to discuss the highlights (and lowlights) of the past week, and plan for the week ahead. The week we met, Fractal was hiring and Velamakanni reviewed with the head of human resources the hiring plans, as well as the training programme for the new hires. The entrepreneur, who believes in "management by walking" (he walks up to the staff instead of summoning them), then met the head of Fractal Sciences for a discussion on the customer genomics algorithm. "It's an algorithm we are designing that tries to figure out what kind of person you are and what are the kinds of things you would like to buy based on the data obtained from your earlier purchases," says Velamakanni.
Post-lunch the team was busy with a client visit; most of Fractal's customers are based in the US and interact on the phone, but every now and then a client visits the Mumbai office.
With 600 employees spread over seven locations, and a bulk of the customers based in the US, Velamakanni says he is always on a plane. Most interesting project: "So many. We started our business with designing a model for ICICI that would reduce their loan default rate. In 2001, we did a project for Hindustan Unilever and later one for Citibank, where we built a model that sold personal loans on a credit card for them, which they loved," says Velamakanni.
Biggest challenge: "In 2009-10, we lost 10% of our revenue in a single day when General Motors went bankrupt. Earlier in 2007-08, we lost 25% of our team in one day. They wanted to start another organization but once funding looked dim, they joined another organization. That was the toughest time for me personally," he says.
Skill set needed for this field: "Problem-solving abilities. Humility because, at the end of the day, ours is a service business. Basic learnability. People with economics and statistics background, because we are building econometric models. Training in computer science helps in building machine-learning models. People with sociology, psychology or anthropology background that can help us try to understand the human mind," he says.
Money matters: The starting salary can be around Rs.7.5 lakh per annum. Data professionals in the US earn upwards of $1 million (around Rs.6.2 crore). In India, at the CEO level, it is upwards of Rs.1 crore.
Every month, we explore a profession through the lives of three executives at different stages in their careers.