https://github.com/razamehar/telco-customer-churn-prediction
https://github.com/razamehar/telco-customer-churn-prediction
Last synced: 4 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/razamehar/telco-customer-churn-prediction
- Owner: razamehar
- Created: 2025-06-23T00:29:53.000Z (4 months ago)
- Default Branch: main
- Last Pushed: 2025-06-23T19:02:37.000Z (4 months ago)
- Last Synced: 2025-06-23T20:20:36.350Z (4 months ago)
- Language: Jupyter Notebook
- Size: 5.87 MB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Telco Customer Churn Prediction
This repository contains an end-to-end **MLOps project** built as a final assignment for the [MLOps Zoomcamp](https://github.com/DataTalksClub/mlops-zoomcamp) organized by **DataTalksClub** (site: https://datatalks.club/) and taught by **Alexey Grigorev**.
The project focuses on predicting customer churn for a telecom company using ML pipelines, data versioning, experiment tracking, and model deployment via FastAPI.
## Features
- Data ingestion from Kaggle's Telco Customer Churn dataset
- Data preprocessing (type conversion, NA handling, encoding)
- Train-test split and feature engineering
- Model training using Random Forest with MLflow logging
- Model evaluation with accuracy and precision metrics
- Model persistence with joblib
- Testing using `pytest`
- Code linting using `pylint`
- API deployment using FastAPI
- Experiment tracking using MLflow + DagsHub
- Feature scaling using StandardScaler
- Data versioning using DVC
- Docker containerization for consistent reproducibility
## Example FastAPI Usage
### Root Endpoint
``` http
GET /
```
Response
``` json
{ "message": "This is the Telco Customer Churn Prediction API" }
```
### Predict Endpoint
``` http
POST /predict
```
Sample Request
``` json
{
"gender": "Female",
"tenure": 12,
"MonthlyCharges": 80.5,
"TotalCharges": 956.2
}
```
Response
``` json
{
"prediction": 1
}
```
## Tech Stack
- **Python 3.10**
- **Data & Modeling**: pandas, scikit-learn, joblib
- **Experiment Tracking**: MLflow, DagsHub
- **API & Deployment**: FastAPI
- **Containerization**: Docker
- **Data Ingestion**: Kaggle (via kaggle CLI or KaggleHub)
- **Version Control**: DVC (Data Version Control)
- **Testing & Code Quality**: pytest, pylint
## MLflow Experiment Tracking
- Tracking URI: DagsHub MLflow Dashboard (https://dagshub.com/razamehar/telco-customer-churn-prediction.mlflow)
- Logged artifacts:
- Parameters: max_depth
- Metrics: accuracy, precision
- Artifacts: trained model (.pkl), scaler (.pkl)
## Run Locally
1. Clone the repository
```bash
git clone https://github.com/razamehar/telco-customer-churn-prediction.git
cd telco-customer-churn-prediction
```
2. Install dependencies
```bash
pip install -r requirements.txt
```
3. Start FastAPI app
```bash
uvicorn app.main:app --reload
```
## Docker Support
### Build Docker Image
```bash
docker build -t churn-predictor .
```
### Run Docker Container
```bash
docker run -d -p 8000:8000 churn-predictor
```
## Run Tests
```bash
pytest tests/
```
## Run Linting
```bash
pylint app/*.py
```
## Contact
For any questions or clarifications, please contact Raza Mehar at [raza.mehar@gmail.com](mailto:raza.mehar@gmail.com).