A full-stack Django web application that transforms raw e-commerce transaction data into actionable business insights through customer segmentation, market basket analysis, interactive dashboards, and AI-powered sales forecasting.
This project is an AI-powered e-commerce analytics platform developed using Django and Python.
The system enables businesses to analyze customer purchasing behavior, identify valuable customer segments, discover frequently purchased product combinations, and forecast future sales using statistical time-series modeling.
The application combines traditional web development with data science techniques to support business decision-making through interactive dashboards and visual analytics.
- Customer segmentation using Recency, Frequency and Monetary (RFM) Analysis
- Automatic customer classification
- Champions
- Loyal Customers
- At Risk
- Can't Lose Them
- Needs Attention
- Lost Customers
- Recent Customers
- Interactive charts and visualizations
- Customer segment distribution
- RFM Matrix visualization
- Revenue insights
- Total Orders
- Total Revenue
- Units Sold
- Average Order Value
- Top Selling Products
- Revenue by Category
- Revenue by Country
- Order Status Distribution
- Interactive charts
Implemented the Apriori Algorithm from scratch without relying on built-in association rule libraries.
Features include:
- Frequent Itemset Generation
- Association Rule Mining
- Support
- Confidence
- Lift
- Configurable Minimum Support
- Configurable Minimum Confidence
- Rule Filtering
Used to discover products frequently purchased together for product recommendations and cross-selling.
Implemented an ARIMA-based forecasting engine for predicting future sales.
Features:
- Automatic Stationarity Testing (ADF Test)
- ARIMA Grid Search
- Automatic Model Selection
- Lowest AIC Selection
- Historical vs Forecast Visualization
- 95% Confidence Interval
- Performance Evaluation
Evaluation Metrics
- RMSE
- MAPE
- Forecast Accuracy
- AIC
Current Model Performance
- Forecast Accuracy: ~85%
- ARIMA Order: (3,1,3)
- Python
- Django
- Django ORM
- HTML5
- CSS3
- Bootstrap
- JavaScript
- Chart.js
- SQLite
- Pandas
- NumPy
- Matplotlib
- Statsmodels
- mlxtend
- Scikit-learn
- RFM Analysis
- Apriori Algorithm
- Association Rule Mining
- ARIMA Forecasting
- Augmented Dickey-Fuller (ADF) Test
- Grid Search Optimization
ecommerce_analytics/
│
├── analytics/
├── customers/
├── forecasting/
├── products/
├── reports/
├── sales/
├── static/
├── templates/
├── screenshots/
├── manage.py
├── requirements.txt
└── README.md
Clone the repository
git clone https://github.com/Paras51Kus/ecommerce-analytics.gitMove into the project
cd ecommerce-analyticsCreate virtual environment
python -m venv myenvActivate virtual environment
Windows
myenv\Scripts\activateInstall dependencies
pip install -r requirements.txtRun migrations
python manage.py migrateStart development server
python manage.py runserverOpen
http://127.0.0.1:8000/
✔ Customer Analytics
✔ Order Analytics
✔ Market Basket Analysis
✔ Sales Forecasting
✔ Interactive Charts
✔ Customer Segmentation
✔ Product Association Rules
✔ Forecast Performance Metrics
- Identify high-value customers
- Improve customer retention strategies
- Discover product associations
- Increase cross-selling opportunities
- Forecast future sales
- Support inventory planning
- Improve business decision making through analytics
- Prophet Forecasting Model
- LSTM Deep Learning Forecasting
- Product Recommendation System
- Real-time Analytics
- Customer Lifetime Value Prediction
- Sales Anomaly Detection
- Power BI Integration
- PostgreSQL Deployment
- REST API Support
Paras Kumar Kushwaha
📧 Email: paraskushwaha891@gmail.com
💼 LinkedIn: https://www.linkedin.com/in/paras-kr-kushwaha-362131314/
🌐 GitHub: https://github.com/Paras51Kus
If you found this project useful, consider giving it a ⭐ Star on GitHub.