AI-Powered Matchmaking Platform for Exporters & Importers 🌍🚀 Welcome to the B2B Global Trade Matchmaking Platform! This project is an advanced, production-ready full-stack application that leverages Artificial Intelligence to seamlessly connect global exporters with matching importers based on comprehensive data analytics, geospatial awareness, and market intelligence.
📖 Table of Contents
Project Overview
Key Features
Architecture & Tech Stack
Project Structure
Installation & Local Development
Environment Variables
Deployment 🌟 Project Overview This application optimizes the international trade supply chain by intelligently matching exporters with prospective buyers (importers). Built with a sophisticated split architecture, it utilizes a Python-based Machine Learning service to rank compatibility using factors such as:
Tariff impacts & Currency shifts Natural calamity risks & Market impacts Intent scoring & Bayesian historical weights It also features a robust, responsive React dashboard that provides users with actionable market intelligence, visual data analytics, and interactive global maps.
✨ Key Features 🧠 AI-Driven Matchmaking Engine: A robust Python backend utilizing scikit-learn, pandas, and Bayesian weighting to score and rank export/import matches intelligently based on industry, geographic multipliers, and risk factors. 🗺️ Geospatial Maps Integration: Interactive mapping powered by Leaflet (react-leaflet) to visualize trade routes, exporter/importer locations, and regional market intelligence. 📊 Intuitive Analytics Dashboard: A comprehensive, white-themed dashboard built with Recharts for real-time visualization of daily, weekly, and yearly trade metrics. ⚡ Real-Time Exploratory Tools: Features like "Explore Importers" and "Export Matches," equipped with interactive modals to seamlessly connect with potential trading partners. 🔒 Role-Based Access & Authentication: Dedicated interfaces for both Exporters and Importers to manage their profiles, recent interactions, and pending deals. 📄 ATS-Friendly Code & UX: Designed with best practices in mind, showcasing modern architectural decisions, clean code standards, and seamless user experiences. 🛠️ Architecture & Tech Stack This project follows a modern Split Architecture pattern:
Frontend (Client) Framework: React 18 (SPA approach with React Router 6) Build Tool: Vite Language: TypeScript Styling: TailwindCSS 3 + Radix UI components + Lucide Icons Data & Maps: TanStack Query (React Query), Recharts, React-Leaflet API Gateway (Server) Runtime: Node.js Framework: Express.js Database ORM: Mongoose / MongoDB Purpose: Manages authentication, user data, data marshaling, and acts as a gateway to the ML service. Machine Learning Service (Python) Language: Python 3 Libraries: pandas, numpy, scikit-learn Purpose: Executes advanced matching pipelines (backend.py, ml_weights.py, feature_engineering.py), applies MinMaxScaler, and processes PCA & Bayesian weights. 📁 Project Structure text Epsilon_LOC8AI1/ ├── client/ # React SPA frontend │ ├── components/ # Reusable UI elements (Radix + Tailwind) │ ├── pages/ # Application views (Dashboard, Login, Explore, etc.) │ ├── App.tsx # React Router configuration │ └── global.css # Global Tailwind styles ├── server/ # Express API Node.js Gateway & Python ML Service │ ├── routes/ # API routes │ ├── backend.py # Python Matchmaking logic │ ├── ml_service.py # ML API Gateway hooks │ ├── data_cleaning.py # Python Preprocessing scripts │ └── index.ts # Node API entrypoint ├── shared/ # Shared TypeScript interfaces & types └── package.json # Node dependencies and npm scripts 💻 Installation & Local Development Prerequisites Node.js (v18+ recommended) Python 3 PNPM or NPM Setup Instructions Install Node Dependencies:
bash pnpm install
npm install Install Python Dependencies: Navigate into the project root and install the required ML packages:
bash pip install -r server/requirements.txt (We recommend using a virtual environment: python -m venv venv)
Start the Development Servers: The project is heavily integrated using Vite and concurrently runs the backend and frontend.
bash npm run dev This command will start the Vite frontend server and trigger the backend data integrations. Note: Verify any Python microservices scripts (e.g., uvicorn or flask if exposed manually) are running if testing ML features in isolation.
🔑 Environment Variables Create a .env file in the root directory and ensure the necessary keys are configured (do not commit your .env!). The system generally expects:
env PORT=8080
MONGO_URI=mongodb+srv://:@cluster.xyz.mongodb.net/
🚀 Deployment The unified system is designed for flexibility across modern cloud providers, specifically tailored for separate host environments given the split architecture.
Frontend: Optimized for deployment on platforms like Vercel or Netlify. Run npm run build:client to generate the static files. Backend (Node & Python): Optimized for deployment on platforms like Render or Railway. The Procfile and package.json build scripts specifically cater to installing both Node and Python requirements in the cloud. 🤝 Contributing Contributions, issues, and feature requests are welcome! Feel free to check out the
issues page to see upcoming features or to report bugs.
Built with ❤️ for a more connected global trade ecosystem.