Mini Student-Advisory Chatbot System: Backend Documentation

Project Overview

The Mini Student-Advisory Chatbot System is a specialized backend application designed to facilitate study-abroad consultations via an AI-driven interface. This system provides a secure and scalable suite of REST APIs for user management, real-time chatbot interactions, conversation persistence, and administrative oversight.

The architecture is modular, production-ready, and designed for seamless integration with any modern frontend framework.

Core Objectives

• Intelligent Query Handling: Process student inquiries using a Large Language Model (LLM).
• Data Persistence: Maintain comprehensive logs of user interactions and conversation history.
• Security & Scalability: Implement robust authentication, rate limiting, and administrative controls.
• Deployment Readiness: Support containerized environments via Docker.

Technical Stack

Layer	Technology
Runtime Environment	Node.js
Web Framework	Express.js
Database	MongoDB with Mongoose ODM
Authentication	JSON Web Tokens (JWT) & bcrypt
AI Engine	Google Gemini API
Infrastructure	Docker, Docker Compose

Project Structure

src/
├── app.js                  # Express application configuration
├── server.js               # Entry point
├── config/
│   └── db.js               # Database connection logic
├── controllers/            # Request handling logic
│   ├── auth.controller.js
│   ├── user.controller.js
│   ├── chat.controller.js
│   ├── conversation.controller.js
│   └── admin.controller.js
├── middlewares/            # Request interceptors
│   ├── auth.middleware.js
│   ├── admin.middleware.js
│   └── rateLimit.middleware.js
├── models/                 # Mongoose schemas
│   ├── User.js
│   ├── Query.js
│   └── Conversation.js
├── routes/                 # API route definitions
├── services/
│   └── llm.service.js      # Gemini API integration
├── Dockerfile
├── docker-compose.yml
└── .env

Configuration

Create a .env file in the root directory with the following variables:

PORT=5000
MONGO_URI=mongodb://localhost:27017/chatbot
JWT_SECRET=your_secure_jwt_secret
JWT_EXPIRES_IN=7d
GEMINI_API_KEY=your_gemini_api_key

Note: Ensure .env is included in your .gitignore to prevent sensitive data leaks.

Installation and Deployment

Local Development

1. Install dependencies: npm install
2. Start the server: npm run start
3. Access the API at: http://localhost:5000

Docker Deployment

To spin up the entire stack (App + Database):

docker-compose up --build

API Reference

All endpoints can be tested using Postman or any REST client. Protected routes require the JWT token to be passed as: Authorization: Bearer

Authentication

• Login
  • POST /api/auth/login
  • Payload: {"email": "user@example.com", "password": "password123"}
  • Returns: {"token": "JWT_TOKEN"}

User Management

• Register User
  • POST /api/users
  • Payload: {"name": "John Doe", "email": "john@example.com", "password": "password123"}
• Get All Users (Protected)
  • GET /api/users
• Get User by ID (Protected)
  • GET /api/users/:id

Chatbot Interaction

• Submit Query
  • POST /api/chat
  • Rate Limit: 20 requests per 15 minutes.
  • Payload: {"userId": "ID", "question": "Query text"}
  • Logic: Validates user -> Calls Gemini LLM -> Stores Conversation -> Returns Response.

Conversation History

• User History: GET /api/conversations/user/:userId
• Recent Activity: GET /api/conversations/recent

Administrative Endpoints

Requires valid JWT and Admin role.

• User Audit: GET /api/admin/users
• Query Logs: GET /api/admin/queries
• Global Conversations: GET /api/admin/conversations

Database Schema Design

User Collection

• name: String
• email: String (Unique)
• password: String (Hashed)
• phone number:String(optional)
• role: Enum (user, admin)
• createdAt: Date

Query Collection

• user: ObjectId (Ref: User)
• question: String
• status: Enum (pending, answered, failed)
• llmProvider: String

Conversation Collection

• user: ObjectId (Ref: User)
• query: ObjectId (Ref: Query)
• question: String
• response: String
• createdAt, updatedAt: Date (auto-generated)

Security Implementation

• Authentication: Stateless JWT-based auth.
• Encryption: Passwords hashed using bcrypt with a salt factor of 10.
• Traffic Control: express-rate-limit applied to AI endpoints to prevent API abuse.
• Architecture: Separation of concerns using the Controller-Service-Repository pattern.

System Health

• Check Status: GET /health
• Response: {"status": "ok", "message": "Server is healthy"}

Screenshots

• /screenshots/

Deliverables

This submission includes:

1. Complete backend source code hosted on GitHub
2. README documentation covering:
   • Project overview
   • Setup instructions
   • Environment variables
   • API usage examples
   • Tech stack details
3. Complete Postman Collection

Acknowledgments

Special thanks to GlobeTrek Overseas for this opportunity.