Decision of loan approval could take lots of efforts because there are various factors to consider during the approval process. And there is no sure way or determined method to make an approval decision.

What it does

I am building a loan application solution which use TypingDNA for authentication and an AI classification model to provide instant loan approval decisions. The model is built based on a set of features derived from some loan data of SME (small and medium enterprises) in Malaysia.

How I built it

First I used the TypingDNA API in the project. I built the solution based on one of the samples of TypingDNA's tutorial. Then I used Jypiter notebook to build an AI model. With some company data and their loan status, I trained a simple model at Kaggle platform and host the API at Heroku.

Challenges I ran into

I did not encounter much problem making the project. I can find all the help from the Internet.

Accomplishments that I'm proud of

I managed to extend the TypingDNA tutorial with a loan application form calling the AI model REST API.

What I learned

In this project I learned the new TypingDNAs' typing biometrics authentication technology. And I too learn how to build an AI model using Kaggle platform.

What's next for Automated loan onboarding

The next step will be getting more and better data to build a real life loan application solution.

Built With

Share this project: