Customer Churn Intelligence Platform

📌 Overview

This project is an end-to-end data science solution designed to predict customer churn and provide actionable business insights. It combines data analysis, machine learning, and deployment into an interactive web application.

🚀 Key Features

• Exploratory Data Analysis (EDA) with business insights
• Feature engineering and preprocessing pipeline
• Machine learning models (Logistic Regression & Random Forest)
• Model evaluation with focus on business impact (churn recall)
• Interactive Streamlit web application

📊 Dataset

Telco Customer Churn dataset (~7,000 customers)

🧠 Key Insights

• Customers on month-to-month contracts have significantly higher churn rates
• Higher monthly charges correlate with increased churn
• Contract type, tenure, and pricing are the strongest churn drivers

🤖 Model Performance

Model	Accuracy	Churn Recall
Logistic Regression	0.80	0.57
Random Forest	0.79	0.50

Logistic Regression was selected as the final model due to better performance in detecting churned customers.

🖥️ Application

The application allows users to:

• Input customer data
• Predict churn probability
• Understand risk level instantly

⚙️ Tech Stack

• Python (Pandas, NumPy, Scikit-learn)
• Streamlit
• Matplotlib / Seaborn
• Git & GitHub

📁 Project Structure

customer-churn-intelligence-platform/ │ ├── data/ ├── notebooks/ ├── models/ ├── app/ ├── reports/ └── README.md

🎯 Business Impact

This project demonstrates how machine learning can support customer retention strategies by identifying high-risk customers and enabling proactive intervention.

---

👤 Author

Trpo Stojkoski

• GitHub: https://github.com/StojkoskiT
• LinkedIn: https://www.linkedin.com/in/trpo-stojkoski/