Help students stop Bank Fraud in India

I wonder why Standard Chartered Bank takes YEARS to analyse and detect fraud credit card transactions.

23-YR-OLD SAKTHIVEL’S PROGRAMME ANALYSES DATA AS IT COMES — UNLIKE THE END-OF-THE-DAY EVALUATION THAT BANKS CURRENTLY USE

College student devises new programme to detect banking fraud
Sreejiraj Eluvangal (from Express Newsline)

Lucknow, May 17: HE may soon turn out to be the hottest property in the ‘banking solutions’ field of the country’s booming IT sector, but 23-year-old Sakthivel is more worried about getting a job right now. The final year student from a small- time college in Tamil Nadu’s Thanjavur district may have found the perfect cure for the rising instances of banking-related crimes.

‘‘If you use my system, you will be able to cut down the time required to detect suspicious bank transactions from hours to just minutes,’’ he says, standing on the sidelines at a technology seminar in Lucknow.

Advertisement
Citibank
Developed as part of his final-year project, Sakthivel’s approach to fraud detection is based on real-time analysis of transaction data — unlike the end-of-the-day analysis currently opted by banks in India.

‘‘If you can detect a suspicious transaction as soon as it happens instead of scouring for it at the end of the day, the chances of preventing further mischief are much higher,’’ he says.

Though real-time fraud detection currently suffers from the risk of bringing the entire network down, Sakthivel has used the latest software tools to get over it.

‘‘The last thing banks want is more work for their transaction servers which process customer requests from ATMs etc. But this approach puts them at the risk of crashing in case of overuse,’’ he explains.

A new software tool called ‘Aglets’ — developed by IBM’s Tokyo centre — helped Shathivel.

Tiny self-contained units of code, ‘Aglets’ collect information from designated computers on their network, dock onto a computer and uses its processing power to process the data it has collected. ‘‘In fact, even if the computer they are lodged in becomes busy these units have the ability to automatically migrate to another one,’’ he says. He has been working on the project for a year at the headquarters of ‘Polaris software’ in Chennai.

While some experts supported Shakti’s claims, others were sceptical.

‘‘We have worked for nearly 10 years with a team of 100 to develop our product,’’ said one of the engineers from a leading software vendor. ‘‘Suddenly when someone comes with such claims, it has to be taken with a pinch of salt,’’ he added. T Senthilkumar, senior lecturer of Computer Engineering at the ’Amrita Vishwa Vidya Peetam’ at Coimbatore and one of the experts consulted by the student, however, was optimistic. ‘‘The technology is a relatively untapped one and from what he told me about the project, I think it is quite a promising one. Of course, in this field, it sometimes takes more time to refine a product than to develop the first version,’’ he said.

Shaktivel has already done his initial checking. ‘‘It has been stable to the extend of a few thousand transactions in two hours and the average time taken to spot rogue transactions was 2 minutes!’’ Of course, it is a long way from the tens of lakhs of transactions per 12 hours that such systems are used to in real life, but I am sure it will take it,’’ he says.