Shloak Rathod built Hedgehog, an automated loan approval system that decides applications in minutes with 88% accuracy. It learns from past loans to flag fraud and check legal compliance, cutting the need for human reviewers. The plan is to add more data and switch to reinforcement learning to handle real-world lending better.
Introducing Hedgehog, an automated online loan approval service. Get your loans approved within minutes. Advantage Hedgehog. Our model aims to learn from what we deem as good loans and bad loans, considering our lending capacity and our loan analytics. We also aim to automate legal compliances and add fraud detection systems to make our models better. Lastly, lastly, the automated service will help us cut down our human resource costs too. So that's our project demo. This is our MVP. So, uh, we just put in our preliminary data. So I'm just going to put in my data. Uh, in the future, we aim to add more data points, um, and therefore better evaluate, um the loan seeker. Let's, um, yeah, and in the future, I'm also, uh, we also plan to automate the credit rating score service so it does not have to be a manual input. And let me submit my application. So the model has given me a result, um, saying that we do not qualify for this loan, and that is rightly so. However, in the future, we aim to implement a reinforcement learning-based model. Currently, our model is a supervised machine learning model which did achieve an accuracy of 88% on its dataset. However, that may be limited in a real-world situation. So, thank you.
