🛡️ AI-Powered Customer Churn Prediction Dashboard

An intelligent, interactive dashboard built with Streamlit + Plotly
for analyzing, visualizing, and predicting customer churn with real-time insights.

🌐 Live Demo | 🔗 GitHub | 👨‍💻 Author

---

🚀 Overview

The Churn Prediction Dashboard is a machine learning-powered web app that identifies which customers are at risk of leaving a service. Built with real-world telecom data, this tool helps business teams proactively engage with users before churn happens.

It features:

• Interactive churn prediction
• Real-time risk recommendations
• Gender-wise and contract-wise visual insights
• Future scenario simulation
• CSV download support

---

🧩 Features

✅ Predict churn probability per customer (0 = Active, 1 = Churned)
📈 Visualize top churn drivers using feature importance
🔍 Smart charts: Tenure, Monthly Charges, Gender Analysis
🔮 Future Churn Simulator (change charges & tenure)
📥 Download full customer dataset
🌑 Sleek dark-mode dashboard UI
🧠 Trained with XGBoost ML Model

---

🛠 Tech Stack Used

This project combines modern ML tools with a rich frontend:

• 📊 Streamlit – For interactive dashboard UI
• 📈 Plotly & Seaborn – Visualizing customer trends
• 🐍 Python – Core logic and modeling
• 🧮 Pandas & NumPy – Data preprocessing & handling
• 🧠 Scikit-learn & XGBoost – Model training and evaluation
• 📄 CSV Dataset – Kaggle’s Telco Customer Churn Dataset

⚙️ Deployed on Streamlit Cloud.

---

🧠 Machine Learning Model

We use XGBoost as the main model:

• Trained on engineered features like tenure, contract, monthly charges, etc.
• High accuracy in predicting churn likelihood
• Feature importance extracted for business interpretation

---

📸 Visual Insights

---

📁 Dataset Columns

We use the following key columns after cleaning and encoding:

gender, tenure, MonthlyCharges, TotalCharges, Contract, InternetService, PaymentMethod, Churn

▶️ Getting Started

Ready to launch the AI Sales Forecasting Dashboard on your machine? Follow these simple steps:

---

🔧 Setup

git clone https://github.com/anshkumar2311/AI-Powered-Churn-Prediction.git
cd AI-Powered-Churn-Prediction
pip install -r requirements.txt
streamlit run app.py

---

📥 Download Filtered Data

🎯 Purpose: Empower users to understand, analyze, and extract customer behavior and churn risk — instantly!

The dashboard offers an intuitive and seamless way to:

• 🧠 AI-Powered Risk Scores per customer
• 📊 Top 10 Sample Customer Profiles with:
  • Tenure
  • Charges
  • Senior Citizen status
  • Actual Churn status
  • Predicted Risk Level

Once explored, users can easily:

✅ Download the full churn dataset as CSV
📤 Use it for reporting, analysis, or business decisions

---

🌟 Show Some Love

If you like this project, please consider:

⭐️ Starring the repo

Let’s connect and grow together 🚀