Loan Eligibility Checker

Built during a hackathon, this project explores how machine learning can make loan decisions more transparent and user-friendly. New founders often don’t know what factors affect lending outcomes, so we designed a lightweight tool that turns a few key inputs into an easy-to-understand risk/eligibility estimate. I developed a Streamlit interface that guides users through entering organization and loan details (industry, state, employee count, loan amount, loan term, and whether the business is new). The app runs an end-to-end ML pipeline (preprocessing + RandomForest classifier) and returns a probability-based score, then generates a readable summary and a downloadable text report. Beyond prediction, the project includes a small data exploration section to visualize patterns (e.g., loan frequencies by state). The final result is a functional prototype demonstrating the full ML workflow-from data prep and model training to deployment-style inference inside a usable web UI-while also emphasizing privacy (no user input persistence) and responsible interpretation of model outputs.