Telco Customer Churn Prediction

This project provides a web interface for predicting customer churn using a Logistic Regression model. It is configured for deployment on Vercel.

Structure

• api/index.py: The Flask backend that serves the model prediction endpoint.
• public/index.html: The frontend UI.
• train_model.py: Script to train the model and save it as model.pkl.
• model.pkl: The trained model.
• schema.pkl: Schema of the features used.

Local Setup

1. Install dependencies:

   pip install -r requirements.txt

2. Train the model (if model.pkl is missing or you want to retrain):

   python train_model.py

3. Run the API locally:

   python api/index.py

   Open http://localhost:5000 in your browser.

Deployment to Vercel

1. Push this repository to GitHub/GitLab/Bitbucket.
2. Import the project in Vercel.
3. Vercel will automatically detect the configuration in vercel.json and deploy the Python serverless function.
4. Ensure model.pkl and schema.pkl are included in the repository .

API Usage

Endpoint: POST /api/predict

Payload example:

{
    "Gender": "Male",
    "Senior Citizen": "No",
    "Partner": "No",
    "Dependents": "No",
    "Tenure Months": 12,
    "Phone Service": "Yes",
    "Multiple Lines": "No",
    "Internet Service": "Fiber optic",
    "Online Security": "No",
    "Online Backup": "No",
    "Device Protection": "No",
    "Tech Support": "No",
    "Streaming TV": "No",
    "Streaming Movies": "No",
    "Contract": "Month-to-month",
    "Paperless Billing": "Yes",
    "Payment Method": "Electronic check",
    "Monthly Charges": 70.0,
    "Total Charges": 840.0
}