In front of a frenzied crowd
in the flood-lit Wanderers Stadium in Johannesburg, Royal Challengers captain
Anil Kumble made a bold move. In the final match of the second season of the
Indian Premier League (IPL) Twenty20 Championship, Kumble, a leg spinner, bowled
the first ball of the match, a task normally undertaken by a fast bowler.
And when he uprooted the wicket of star batsman Adam Gilchrist in the third
ball of the first over, what seemed like a bold gamble on his part had paid off.
But Kumble, himself an engineer, relied on more than just instinct. He had roped
in a small start-up in Bangalore, Sportingmindz, to analyse Gilchrist's batting
style. Using data analytics technology, the start-up had honed in on the
Australian batsman's achilles heel - spin bowling. "From then on, it became a
trend to attack Gilchrist with spin bowlers," said Sandeep Kannambadi,
co-founder of Sportingmindz.
From developing strategies for cricket
teams to detecting spurious drugs and even predicting a crime, India data
scientists are building specialised systems that can chew through billions of
bits of data, analyse them via self-learning algorithms and package the insights
for immediate use.
CRICKET'S BIG
DATA Sportingmindz
has captured and collected terabytes of cricket videos and data in its central
server, enough to store millions of books, photographs and songs. Customers
include South African, New Zealand and IPL cricket teams and the information is
kept strictly confidential.
"Surprisingly, one of our data scientists
is even travelling to Gulbarga in Karnataka to do analytics for a school cricket
team, who is also our client," said Kannambadi, who added that Microsoft's
start-up platform BizSpark helped them to get software tools and connected with
industry players. Kannambadi, along with Sanjay Rao, quit their jobs at India's
third-biggest software services provider
Wipro
in 2006 and joined hands with former Indian cricketer Vijay R Bharadwaj to start
Sportingmindz. Besides cricket teams, the start-up is tapping other sports as
well, such as hockey, tennis and table tennis.
PREDICTING CRIME Crime fighting has also found an ally in big
data analytics. The Tom Cruisestarrer 'Minority Report', set in 2054, showed
prediction of crime. Data scientists at the India laboratories of
IBM
have brought that future to the present. "We predict the chances of a person
committing a crime, based on jail records," said Vishwanath Narayan, IBM's chief
technology officer for Industry Solutions Architecture. IBM India has deployed
this analytics for the police departments of New York and Chicago and is now
working with an Indian state police agency.
IBM also delivers traffic
prediction in cities such as London and Stockholm, for which the prediction
algorithms are done out of India. The traffic prediction is based on real-time
information that helps IBM identify the patterns of how a particular roadway is
getting congested and traffic and transport authorities of these cities are able
to change the pricing of tolls accordingly.
DETECTING SPURIOUS DRUGS Bangalore-based data analytics
startup
Activecubes is
analysing huge chunks of data to detect spurious drugs in the country. The firm
has developed a product where customers can send a unique code, available on the
drug at the time of purchase, by text message to a common number. The message
comes to the Activecubes database, which responds back in a few seconds about
the authenticity of the drug and provides other miscellaneous information.
Activecubes is betting on a huge opportunity in combating counterfeit
drugs, valued at $75 billion worldwide. In India, around 30% of the drugs sold
in the market are spurious. It has pinched the nerves of many pharma companies
whose key brands have lost almost 30% share of the market in India because of
spurious drugs. "My son was not well and the medication was not working. First
thing I told my wife was I don't know whether that was a spurious drug," said
Rajesh Varrier, CEO and co-founder of Activecubes, who is tying up with a few
pharma companies to implement the product.
CUSTOMER PREFERENCE Data analytics is also being used to
understand customer choices. One of world's premium chocolatiers, a renowned
160-year-old Swiss brand, commissioned social media analytics startup Salorix to
analyse conversations on all popular social media channels. Salorix, founded by
Santanu Bhattacharya collected and semantically analysed more than 100,000
conversations, including that of competitor brands, over a period of four weeks.
The chocolatier's brand managers were able to participate in these
conversations, raising the visibility of the brand during a crucial holiday
season.
US-based food and pharmacy retail major Haggen tur ned to
Bangalore-based retail analytics firm Manthan Systems to get a better view of
customer. Their meat section was running promotions on steaks and high-end meat
cuts on the assumption that these meats attract customers. But after deploying
Manthan's technology, the company realised it was in fact the cheaper ground
beef that was drawing in the customers and started running promotions on this
product.
Similarly their in-store sushi deli made fresh sushi every
morning. With the analytics solution, Haggen managers realised that most sushi
sales happened post-5pm. The managers made labour-saving changes like making
sushi only in the afternoon.
Another Bangalore-based data analytics
firm
Mu Sigma
analyses facial expressions of customers, eye gaze and body language across
supermarkets and shopping malls to understand customer behaviour. They crunch
thousands of hours of video footage captured by small cameras embedded at retail
point of sale locations. It then sends the information to the retailers to help
them understand the mood, attitude and actual behaviour of consumers and help
companies to better position their products.
"Planets are aligned.
This is a golden era of analytics, it will give birth to new stars," said Atul
Jalan, founder and CEO of Manthan Systems, which analyses consumer behaviour for
brands like Carrefour,
Louis
Vuitton, Prada and Crocs, and is expecting to touch $100 million revenue by
next year.
Digg
ApnaCircle
