Credit Ease is an attempt to assist credit score generation with Artificial Intelligence.

Using A.I. reduces the chances of unfair results as it predicts score with complex logic which surpasses most human or formula based score generations. It is able to find patterns, classify and even recommend suggestions to achieve a certain level of credit-worthiness.

We started with a blank canvas and and Idea. We had to find a way to predict score without using conventional techniques of grinding down on one's credit history. We wanted to find the patterns in betweeen the numbers. And hence began our journey.

We dwindled through tons of datasets and chose to select a really dense dataset with input features being only personal and day-to-day information. It was a dataset for credit risk. We began the operation by cleaning and removing unnecessary features which weren't supposed to be passed to our model. After encoding, scaling and all pre-processing, came the real question, which model to use.

We decided to get our hands dirty and tried to optimize and find results for most of the popular classifiers and models. We furthered tweaked and tested the best hyper-parameters for the models which survived last. The winner at the end was K Nearest Neighbours Classifier. Upon further tinkering we had built our finest model.

We also processed the data of n nearest neighbours of our input data, so we could find differences and recommend possible upgrades to the user whih maximizes their chance at better credit worthiness.

With our existing knowledge of web technologies, we built a app built with React.js in client and flask in server. We imported our model and processing functions to server and made API methods to fetch scores and predictions u[on requests from the client-side.

In future, we can work on better datasets and try neural networks in hope of improving and solidifying this concept. We can build smartphone apps , microservices, and provide solutions for the common man