Campus Marketplace

A high-performance, campus-exclusive marketplace platform designed for students to trade textbooks, gadgets, and essentials. This project implements a robust microservices architecture using Go and Next.js, featuring AI-powered search and real-time communication.

🖼️ System Design & Architecture

Diagrams

The platform is built on a distributed microservices architecture to ensure high availability and independent scalability:

🚀 Key Features

• Microservices Architecture: Independently scalable services for orchestration, listings, and real-time events.
• AI-Powered Search: Natural Language Processing via Google Gemini API to handle queries like "looking for a used laptop under $500."
• Real-time Chat: Full-duplex communication using WebSockets for buyer-seller negotiations.
• Presence Tracking: Real-time online/offline status management powered by Redis.
• Reliable Messaging: Asynchronous chat processing via RabbitMQ with persistent message status (Undelivered, Delivered, Read).
• Listing Management: Complete CRUD operations with cloud-based media storage via Azure Blob Storage.

🛠 Technology Stack

Backend (Go)

• Go 1.23+: Core backend services (Orchestrator, Events, Listing).
• PostgreSQL 16: Relational storage for users and listings.
• MongoDB 7: Document-based storage for chat history.
• Redis 7: High-speed presence and message delivery tracking.
• RabbitMQ 3: Message broker for decoupled service communication.
• Google Gemini API: LLM integration for intelligent search parameters.

Frontend (Next.js)

• Next.js 16 / React 19: Modern SSR framework.
• TypeScript: Type-safe frontend architecture.
• Tailwind CSS 4: Utility-first styling.
• WebSocket API: Bi-directional event handling for real-time features.

🏗 System Components

1. Orchestrator Service

The central API gateway handling authentication (JWT), session management, and request routing. It coordinates data flow between the frontend and internal services.

2. Events Server & Chat Consumer

Handles real-time logic. When a message is sent via WebSocket, it is published to RabbitMQ. The Chat Consumer then checks Redis for recipient presence:

• Online: Delivers the message immediately.
• Offline: Persists the message to MongoDB marked as "Undelivered" for later synchronization.

3. Listing Service

Manages the marketplace engine. It transforms natural language queries into structured database filters using the Gemini API and manages media uploads to Azure Blob Storage.

📁 Project Structure

.
├── orchestrator/      # Central API gateway & Auth
├── listing-service/   # Listing & Search logic
├── events-server/     # WebSocket management
├── chat-consumer/     # Asynchronous message worker
├── frontend/          # Next.js web application
├── http-lib/          # Shared internal utilities
└── docker-compose.yml # Full-stack container orchestration

🚥 Quick Start

Setup

1. Clone the repository

git clone <your-repo-url>
cd campus-marketplace

2. Environment Configuration Copy .env.example in each service directory and the root directory to .env. Provide your credentials for PostgreSQL, MongoDB, Redis, RabbitMQ, and the Gemini API.
3. Deploy with Docker

docker-compose up -d

4. Access Points

• Web UI: http://localhost:3000
• API Gateway: http://localhost:8080

📄 License

Distributed under the MIT License. See LICENSE for more information.