Customer Retention Intelligence is a deployed interactive analytics application built to turn 7,043 telecom customer records into actionable retention insights.

I designed the project as a modular data product rather than a static dashboard. It combines customer data preparation with dynamic segmentation, a tenure × contract churn risk matrix, churn-driver exploration, priority segment analysis, reported churn reasons, financial exposure metrics, and filtered CSV/Excel exports.

The analysis surfaced a clear contract pattern: observed churn was 45.8% among Month-to-Month customers, compared with 10.7% for One Year and 2.5% for Two Year contracts. These results are presented as observed associations rather than causal claims.

The codebase separates data processing, analytics, visualization, and export logic for maintainability. The application was built with Python, Pandas, Streamlit, Plotly, and openpyxl and deployed as a live Streamlit application.

Live Demo: https://nisa-retention-intelligence.streamlit.app/

Source Code: https://github.com/niisa0/customer-retention-intelligence