Fintech App for Viewing CIBIL Scores, Credit Reports, and Red Flags

Solutions We Offered

We began the work with an initial analysis of the practical limitations, improvement areas, and credit information sources we needed to consider. We then finalized an advanced tech stack and started moving ahead with the implementations.

We used Figma to wireframe the layout first. Figma helps customize multiple app screens. So, we used Figma to chalk out the wireframe and then proceeded with the code.

We developed the backend using Node.js because it’s highly extensible. As for the frontend, we chose React Native because it allows creating complex UIs and adding modifications during runtime.

We next integrated the i2i third-party API service to extract and show financial data on the app. The i2i API offers excellent credit information by considering 50+ parameters like people’s debt-income ratios, employability, bank account statements, etc. So, i2i was the ideal fit for drawing credit information. However, during i2i integration, we encountered several 503 Errors and ‘Null’ value issues. But we tracked and fixed them all.

We also used the CRIF API service to extract CIBIL Scores. CRIF helps find large amounts of credit information about people. But, because we didn’t need all the information, we injected rule-based engines to generate CIBIL Scores around specific financial indices. We also set predefined conditions to show Red Flags under ‘potentially risky’ financial conditions.

As for the database, we used PostgreSQL as it has multiple layers for checking bank accounts, balances, statements, etc. It hardens databases against threats and best meets Fintech database demands.

We used the Cashfree payment gateway to allow paid subscriptions. Once users pay for subscriptions, they can avail of advanced services, such as monthly credit reports, expert advice, etc.

At one of these development stages, our client asked us to insert a pre-built dataset of 25K+ financial entries at short notice. We found the dataset had multiple errors. But, we tactfully modified all errors and inserted a well-structured dataset without impacting the rest of the code.

The final testing was challenging, too, as we had to test the app with actual bank statements and not arbitrary docs. However, we tested every functionality successfully without facing setbacks before hosting the app on the AWS server.