GitHub repo(Do give it a ⭐)
Weblink
Excited to share a redeveloped customer churn prediction web app, now tailored specifically for the banking sector. Originally built for telecom data during a hackathon, this project has been reworked to analyze 4000+ bank customer records and predict churn risk with improved precision.
The app leverages 8 machine learning models, including XGBoost, Random Forest, SVM, and a Voting Classifier, with the highest accuracy reaching 84.25%. It also integrates the Mistral Saba 24B large language model to generate natural language explanations and email drafts for at-risk customers.
A Streamlit-based frontend presents the results through visual insights and user-friendly export options.
Note: Due to OpenAI usage limits, some delays may occur after heavy usage. If you encounter a rate limit, please allow 6–7 minutes before retrying.
Top comments (0)
Some comments may only be visible to logged-in visitors. Sign in to view all comments.