Mihavana/erp_project_e-commerce-madagascar

## Project Team - **RAHOLDINA FIARA ANJARA MIHAVANA** - **ANDRIATSIFERANA NO KANTO LORIDA** - **RAKOTONIRINA MENDRIKA ITOKIANA** --- # Demo Guide: Madagascar E-Commerce Platform (ERP & BI) This project is a comprehensive e-commerce management solution (ERP) coupled with an intelligent Business Intelligence (BI) and Artificial Intelligence stack. ## Project Architecture The system comprises **6 services** orchestrated by Docker: 1. **Database (PostgreSQL)**: Central storage for ERP data and Data Warehouse. 2. **ERP Service (FastAPI)**: Management of orders, customers, and inventory for Madagascar marketplace. 3. **ETL Service (Pandas)**: Automation of extraction and transformation to star schema. 4. **Analytics Service (Scikit-Learn)**: Data Mining engine (RFM, ARIMA). 5. **AI Dashboard (Streamlit)**: User interface with AI-powered reporting (NLG). 6. **Metabase**: Self-service BI tool for visual exploration. --- ## Quick Installation 1. **Set environment variables**: Create a `.env` file at project root: ```env # Database DB_USER=admin DB_PASS=123456 DB_NAME_ERP=erp_db DB_NAME_BI=bi_warehouse # AI Token (Free on console.groq.com) GROQ_TOKEN==your_Groq_token_here ``` 2. Launch the infrastructure: ```bash docker compose up -d --build ``` --- ## Demo Steps ### 1. Master Data Initialization Populate ERP with initial products and customers. - **Action**: Click on localhost - **Tool**: Browser or Postman. ### 2. Generate Bulk Data Generate 50+ historical orders for AI model training. - **Action**: Click on localhost ### 3. Launch ETL Transfer ERP data to Data Warehouse BI (Star Schema). - **Action**: Click on localhost - **Verification**: Check `localhost` for loaded rows. ### 4. Data Analysis & Mining Analytics service now processes Warehouse data. - **RFM Segmentation**: `localhost` - **ARIMA Predictions**: `http: …