Inspiration
Me (Timi) and Eyitayo were going out to a pick-up store and we were stopped by law enforcement agents for being potential cybercrime suspects. It got us thinking a lot about the state of financial security in Nigeria and just around the same time, the FSI Hackathon came along.
What Does It Do
Banking services currently use password authentication to validate users. Our solution improves on this by observing both the users’ typing & in-app navigation rhythm to train a neural network. It then compares the current usage pattern to an established rhythm and scores the user. If the score obtained is below an established benchmark, the user would need extra authentication.
To address the issue of compromised card details, our solution enables the authenticated user to generate virtual cards that gives the user more control over the functionality of the card such as where the card can be used, time of day it can be used, single transaction limits etc. These cards can be invalidated after any number of uses. Finally, the user can link the virtual card to their physical card, thereby imposing all the restrictions of their virtual card on their physical card.
How we built it
We built the android app using android studio and the models using keras in python.
Challenges we ran into
Finding the right dataset. Understanding the FSI Sandbox API
Accomplishments that we're proud of
Implementing all the features that we set out to include.
What we learned
We experienced first hand the pressures of innovation and how to work in a team under stress
What's next for FINN-API
To standardize the API for use with any Mobile Banking application.
Log in or sign up for Devpost to join the conversation.