It was his caffeine habit that gave Mr Vikram Bahl the kick he needed to create CashTrack, an app for managing personal finances.
"Every month, when I received my bank statement, I was aware of the total amount I was spending, but I didn't know exactly what I was spending on," said the 27-year-old software engineer.
It was only after he studied the breakdown of his expenses more closely that he became aware of some of his habits, such as shelling out $500 in six months for his daily cup of Starbucks coffee.
"I think it's human psychology. You may not realise how much you are spending, but when the total is staring you in the face, then you think, 'Okay, maybe I should cut down on the Starbucks'," he said.
In order to make it easier for people to manage their finances, Mr Bahl and his co-developer, research engineer Vasundhara Jayaraman, 27, came up with the idea of CashTrack last September.
The pair met while both were studying for a master's in computer science at the Nanyang Technological University.
The app was named the Most Intuitive Personal Financial Management Solution at the recent Asia-Pacific iteration of the Citi Mobile Challenge, a competition organised by Citibank to develop mobile banking solutions.
The Citi Mobile Challenge was first launched in Latin America in April 2014, and in the Asia-Pacific in August last year. Citibank offers a share of US$100,000 (S$142,000) in cash awards to selected finalists.
The app will be released in two months' time on iOS, Android and Web platforms.
Mr Bahl and Ms Jayaraman have not yet set up a company to market the app, but will be registering one within the next month. They will first launch the app with Citibank, and have plans to expand to other banks at later dates.
The app converts transaction codes to merchant names, offers a categorised breakdown of the type of expenses incurred and provides options for users to invest savings.
After CashTrack users log into the app with their Internet banking credentials, they will be able to see merchant names instead of transaction codes, which can sometimes be difficult to understand.
"For example, one of the transactions on my statement said 'SBUX', which I had to Google to find out was Starbucks,"said Mr Bahl.
"Also, if you eat out, the transaction may be made under the name of the restaurant's parent company, which can be confusing."
CashTrack also automatically breaks down expenses into six categories: utilities, shopping, dining, loans, rental and miscellaneous.
"Bank statements are usually in the form of a list, which is not organised. But this gives you more details about your personal habits, like how much you have spent in the last month on dining or shopping," he said. "For example, if you're spending a lot on food, you may decide to eat out less and cook more."
Both matching the merchant name to the transaction code and categorising expenses are mammoth tasks as there are no ready databases available.
In order to generate such information, Mr Bahl and Ms Jayaraman have been using big data analysis and machine learning algorithms to parse records of bank transactions.
The CashTrack app does not just allow users to keep track of their spending, but also to invest and grow their savings.
Once a user starts saving a certain amount every month, CashTrack will begin showing personalised suggestions on how to invest it.
Mr Bahl said: "For example, if a person has a lot of car loans or housing loans, it will not suggest doing something aggressive like currency trading, but to save up and pay back more of the loan and invest in less risky things like mutual funds or fixed deposits."
As it will be free to use, commission from these investment products is likely to be the main source of revenue for the app, although Mr Bahl said that the finer details of the app's business model have not yet been finalised.
CashTrack has also spurred Mr Bahl to cut down on his coffee habit: "Instead of going to Starbucks every day, I now go about two times a week.
"The rest of the time, I mix instant coffee sachets with hot water."
This article was first published on January 6, 2016.
Get a copy of The Straits Times or go to straitstimes.com for more stories.