Cuana is an end-to-end analytics suite that integrates churn/CLV prediction, customer segmentation and lead scoring into a comprehensive solution using state-of-the-art ML methods, data pipelines and an interactive dashboard. This repository is a fork of my other project initially developed as part of a freelance work published with client approval.
Some features:
- Data pipeline
Clean, preprocess and split customer data using Python (pandas, NumPy). - Churn & CLV prediction
Implements classification models (RM) for predicting churn and estimating customer lifetime value. - Customer segmentation
Uses both k-means and hierarchical clustering to identify distinct customer groups based on behavior and value. - Lead scoring
Builds a supervised model (XGBoost) to score and prioritize new leads for sales outreach. - Interactive dashboard
Provides a dynamic BI dashboard built with Dash and Plotly to visualize metrics, sales funnel analytics and segmentation outcomes.
.
├── README.md
├── .gitignore
├── requirements.txt
├── src/
│ ├── main.py
│ ├── data_pipeline.py
│ ├── churn_model.py implementation
│ ├── segmentation.py
│ ├── lead_scoring.py
│ ├── dashboard.py
│ └── utils.py
└── tests/
├── test_data_pipeline.py
├── test_churn_model.py
├── test_segmentation.py
└── test_lead_scoring.py
Make sure you have Python 3.8+.
- Clone:
git clone [email protected]:avrtt/Cuana-E2E.git cd Cuana-E2E
- Create and activate a venv:
python -m venv venv source venv/bin/activate # Linux venv\Scripts\activate # Windows
- Install the required packages:
pip install -r requirements.txt
Run:
python src/main.pyThis will:
- load and preprocess the synthetic customer data
- train the churn/CLV prediction, segmentation and lead scoring models
- launch an interactive dashboard to explore the results
Unit tests are located in the tests/ folder. You can run them using:
pytest tests/- Enhance model interpretability with SHAP values
- Integrate real-time data streaming into the dashboard
- Deploy the solution as a cloud-based service
MIT