VedaAI AI Powered Question Paper Generator

An intelligent assessment creation platform for educators designed for high scalability using BullMQ. VedaAI uses Google Gemini AI to generate customized question papers from topics lesson notes and configurable parameters all in real time.

---

Tech Stack

Layer	Technology	Purpose
Frontend	Next.js 16, React 19, TypeScript	App router, SSR, component architecture
Styling	Tailwind CSS 4, Radix UI, Lucide Icons	Design system, accessible primitives, iconography
State	Zustand	Global state management for assignments and real-time status
Forms	React Hook Form, Zod	Form validation and schema-driven inputs
Backend	Node.js, Express 5, TypeScript	REST API, file handling, route management
AI Engine	Google Gemini 2.5 Flash	Question paper generation with multimodal support
Queue	BullMQ	Async job processing for AI generation tasks
Database	MongoDB, Mongoose	Persistent storage for assignments and generated papers
Cache	Redis, IORedis	Response caching and BullMQ job queue backend
Real-time	Socket.IO	Live progress updates during paper generation
PDF	react-to-print	Client-side PDF export of generated question papers

---

System Design

Architecture Flow

1. User submits assignment config (topic, marks, difficulty, question types)
         |
2. Express API creates Assignment in MongoDB (status: pending)
         |
3. Job pushed to BullMQ queue (PaperGenerationQueue)
         |
4. Frontend joins Socket.IO room (jobId) for real-time updates
         |
5. BullMQ Worker picks up job, sends prompt to Gemini 2.5 Flash
         |
6. Gemini returns structured JSON (sections, questions, marks)
         |
7. Worker saves paper to MongoDB, emits AI_COMPLETE via Socket.IO
         |
8. Frontend receives paper in real-time, renders formatted output
         |
9. User can download as PDF or regenerate with one click

API Endpoints

Method	Endpoint	Description
POST	/api/generate-paper	Create assignment and queue AI generation
GET	/api/assignments	List all assignments (cached with Redis)
GET	/api/assignments/:id	Get single assignment with paper
DELETE	/api/assignments/:id	Delete assignment
POST	/api/assignments/:id/regenerate	Regenerate paper for existing assignment
GET	/api/health	Health check endpoint

---

Features

Core Features

• AI Question Paper Generation -- Generate complete, structured question papers from any topic using Gemini 2.5 Flash
• Multimodal Input -- Upload lesson notes as images or PDFs to generate context-aware questions
• Configurable Parameters -- Set topic, difficulty, marks, due date, question types (MCQ, Short Answer, Long Answer, etc.)
• Real-time Generation -- Live progress tracking via WebSockets while AI generates the paper
• PDF Export -- Download generated question papers as formatted A4 PDFs
• Regeneration -- One-click regenerate to get a fresh set of questions

Assignment Management

• Assignment Dashboard -- View all assignments in a responsive card grid
• Search and Filter -- Search assignments by topic, sort by newest, oldest, or due date
• CRUD Operations -- Create, view, and delete assignments
• Status Tracking -- Visual status badges (Completed, Generating, Failed)

Pages

• Home -- Dashboard with stats overview (total assignments, completed, in progress), AI hero banner, and recent assignments
• Assignments -- Full assignment management with search, filter, and card grid
• Create Assignment -- Multi-step form with file upload, topic input, date picker, and question type configurator
• Assignment Output -- Rendered question paper with download and regenerate actions
• AI Teacher's Toolkit -- Feature grid showcasing available and upcoming AI tools
• My Groups -- Coming soon preview with feature roadmap cards
• My Library -- Generated papers archive with templates and resources placeholders
• Settings -- Profile info, school details, and preference toggles

UI/UX

• Custom Typography -- Bricolage Grotesque (ExtraBold 800) for headers, DM Sans for body
• Floating Layout -- Sidebar and navbar with rounded corners and margins, creating a modern floating aesthetic
• Responsive Design -- Optimized for desktop and mobile with adaptive card sizes and layouts
• Animated Transitions -- Fade-in animations, hover effects, and micro-interactions
• Toast Notifications -- Sonner-powered feedback for user actions
• Profile Dropdown -- Quick access to GitHub, Portfolio, and LinkedIn links

---

Project Structure

Veda-AI-assignment/
|
|-- Backend/
|   |-- src/
|   |   |-- config/
|   |   |   |-- db.ts              # MongoDB connection
|   |   |-- models/
|   |   |   |-- Assignment.ts      # Mongoose schema (topic, marks, paper, status)
|   |   |-- routes/
|   |   |   |-- api.ts             # REST endpoints + BullMQ queue
|   |   |-- workers/
|   |   |   |-- consumer.ts        # BullMQ worker + Gemini AI integration
|   |   |-- index.ts               # Express + Socket.IO server setup
|   |-- package.json
|   |-- tsconfig.json
|
|-- Frontend/
|   |-- app/
|   |   |-- page.tsx               # Home dashboard
|   |   |-- layout.tsx             # Root layout (fonts, providers)
|   |   |-- assignments/
|   |   |   |-- page.tsx           # Assignment list
|   |   |   |-- create/page.tsx    # Create assignment form
|   |   |   |-- [id]/page.tsx      # Assignment output view
|   |   |-- groups/page.tsx        # Groups (coming soon)
|   |   |-- toolkit/page.tsx       # AI Toolkit feature grid
|   |   |-- library/page.tsx       # My Library
|   |   |-- settings/page.tsx      # Settings page
|   |-- components/
|   |   |-- layout/
|   |   |   |-- AppLayout.tsx      # Main layout wrapper
|   |   |   |-- AppSidebar.tsx     # Navigation sidebar
|   |   |   |-- TopBar.tsx         # Top navigation bar
|   |   |-- ui/                    # Radix-based UI primitives
|   |   |-- EmptyState.tsx         # Reusable empty state component
|   |   |-- providers.tsx          # Theme + query providers
|   |-- store/
|   |   |-- assignmentStore.ts     # Zustand store (assignments, socket, jobs)
|   |-- public/assets/             # Logo, illustrations
|   |-- package.json
|   |-- tailwind.config.ts
|
|-- README.md

---

Getting Started

Prerequisites

• Node.js 18+
• MongoDB (local or Atlas)
• Redis (local or cloud)
• Google Gemini API key

Environment Variables

Backend (Backend/.env):

PORT=8080
MONGODB_URI=mongodb://localhost:27017/vedaai
REDIS_URL=redis://localhost:6379
GEMINI_API_KEY=your_gemini_api_key
FRONTEND_URL=http://localhost:3000

Frontend (Frontend/.env):

NEXT_PUBLIC_API_URL=http://localhost:8080

Installation

# Clone the repository
git clone https://github.com/MIHIR2006/Veda-AI-assignment.git
cd Veda-AI-assignment

# Install backend dependencies
cd Backend
npm install

# Install frontend dependencies
cd ../Frontend
npm install

Running the Application

# Terminal 1: Start the backend
cd Backend
npm run dev

# Terminal 2: Start the frontend
cd Frontend
npm run dev

Open http://localhost:3000 in your browser.

---

How It Works

1. Navigate to Create Assignment from the sidebar or home page
2. Upload lesson notes (optional) -- supports images and PDFs
3. Fill in the topic, due date, and additional instructions
4. Configure question types with count and marks per type
5. Click Generate -- the backend queues a BullMQ job
6. Watch real-time progress as Gemini AI creates your paper
7. Review the formatted question paper on the output page
8. Download as PDF or Regenerate for a fresh version

---

Author

Mihir Goswami

• GitHub: github.com/MIHIR2006
• Portfolio: mihirgoswami.is-a.dev
• LinkedIn: linkedin.com/in/mihir-goswami

---