Inspiration

The concept of Black Boxes is actually derived from the science and engineering fields. Specifically, Black Boxes are defined as something that you can view externally, but have no understanding of how it works internally. The same concept can be applied to credit scores, as many consumers just take note of the three-digit score that they get, yet have no idea of how – and why – it is what it is.

We conducted a survey of the users and collected data from around 50 users, we found the following pain points:

  1. Too many credit score providers
  2. Too many different sets of rules
  3. Unawareness about the factors affecting credit score
  4. Accidental missing of bills and payments
  5. Self-initiative requires every time to know the credit scores
  6. Harsh Penalties

These pain points inspired us to build a solution which will be :

  1. Transparent
  2. Consistent
  3. Reliable
  4. System Initiation
  5. Alleviation
  6. Self Guide
  7. Inclusive of all types of users

What it does

This voice-enabled application uses machine-learning to inform users with particular steps that will help them make informed decisions about their credit usage and purchases and also alarm the user about fraudulent activities using machine-learning.

How we built it

We conducted a survey using Qualtrics and to get the data insights we created dashboards in Tableau Implemented front end prototype using Multiplayer Editing in Figma Used python libraries:

  1. numpy
  2. pandas
  3. sklearn
  4. matplotlib for clustering users according to their demographics and credit scores for fraud detection by finding outliers in the data.

Challenges we ran into

Initially, it was difficult to understand how each transaction of the user affects the credit score and to come up with a solution that addresses this issue turned out to be a major challenge. Upon further research, we decided to build an application prototype integrated with a voice-enabled interface that will inform the user's with specific steps using machine-learning that could help the user improve the credit score and try to demystify the credit score calculation process.

Accomplishments that we're proud of

This prototype when developed into an application will let the users be more informed about the steps that will help maintain or improve the credit score before performing certain purchases.

What we learned

Implementation of Data Science backed by UX Research

What's next for Credible

Share this project:
×

Updates