The problem FINder solves

We are trying to solve a problem that when people meet each other on dating or matrimonial platforms they are matched on fake data. These people have to beleive in the information being conveyed by the other person since there is no way to verify this.

Hence, we come into picture to solve this issue so that people do not end up getting cheated.
We are taking basic details of the user to make their profile and also verifying their financial details through SETU, so that no one can create fake profile and lie about their background or livelihood. We have build an algorithm that identifies the spending patterns of each profile, their monthly earnings and expenditure and their hobbies, interests and activities they like the most and spend time on through the transactions of that person.

We have parsed the transactions of each profile and mapped them to different activities like more spends done at restaurants signify that the person is a foodie or someone making payments from different locations tells us that they travel alot. After we have the hobbies and interests of the user we match them to the profiles with same interests and lie in similar financial circle.

Challenges we ran into

We ran into a couple of challenges of which major one was to identify trends and patterns in setu transaction data since most of the data was limited to a few people and hence their spending paterns were limited. To solve this we went through some of the datasets available on the internet to test our application, and also did a study on identifying and analysing transaction data from different payment gateways.