Making CLAT prep smarter, adaptive, and affordable with AI.
TestVerse is an AI-driven study companion built with Next.js, Prisma, and MongoDB. It replicates the real CLAT exam experience with 120-question passage-rich mocks, adaptive learning, and community features.
🎯 Vision: Democratize access to personalized, exam-accurate CLAT prep with:
- AI-powered test generation
- Adaptive schedules & weak-area tracking
- Gamified, engaging learning experiences
- Secure, scalable, and institute-ready admin tools
- AI Test Engine – Auto-generate CLAT-standard 120Q mocks (online + PDF)
- Adaptive Learning – Personalized schedules, weak-point analysis
- Progress Tracking – Accuracy, study streaks, time analysis
- Gamification – Leaderboards, streak rewards, motivational nudges
- Community Learning – Study rooms, peer sharing, group chat
- 24/7 AI Assistant – Instant doubt-solving with multilingual support
-
User Management
- View all users with profile photo, role, and subscription details
- Role changes (FREE ↔ PAID ↔ ADMIN)
- Block/unblock users in real time
-
Test Creation
- Rich metadata: title, description, type (FREE/PAID)
- Configurable marks, duration, highlights, thumbnail
- Validation + error handling
-
Question Management (in progress)
- Add comprehension, tables, images
- Organize section-wise (English, GK, Legal, Logic, Quant)
- Support multiple Q-types (OPTIONS, INPUT)
-
Advanced Admin Controls (planned)
- Test editing, visibility toggle (active/inactive)
- View test statistics
- Notifications, analytics dashboard, payment integration
- Frontend: Next.js 14 (App Router) + Tailwind CSS
- Backend: Node.js / FastAPI
- Database: MongoDB with Prisma ORM
- Auth: NextAuth.js
- AI/ML:
- Gemini RAG / ChatGPT / HuggingFace APIs
- Semantic Search (Vector DB)
- PDF/Word Scraper for past papers
- Whisper NLP (Hindi-English support)
- Deployment: Vercel / Render + GitHub CI/CD
clatprep/
├── app/
│ ├── admin/ # Admin panel
│ │ ├── users/ # User management
│ │ ├── create-test/ # Test creation
│ │ └── ...
│ ├── api/ # API routes
│ │ └── admin/ # Admin endpoints
│ └── dashboard/ # User dashboard
├── components/
│ └── ui/ # Reusable UI components
├── prisma/
│ └── schema.prisma # Database schema
└── lib/
└── utils.js # Utilities
QuestionType: OPTIONS | INPUTOptionType: SINGLE | MULTISectionType: ENGLISH | GK_CA | LEGAL_REASONING | LOGICAL_REASONING | QUANTITATIVE_TECHNIQUES
- Test: Rich metadata (duration, marks, highlights)
- Question: Complex Q-types with comprehension, tables, images
- Answer: Tracks responses, timing, reporting
- TestAttempt: Total time, attempted count
- User: Role-based + blocking
- Minimal data collected (name, email)
- Data encrypted & anonymized
- Leaderboards show rank only (no personal info)
- No sharing of personal data with third parties
npm installcp .env.example .env.local
# Configure MongoDB + NextAuthnpx prisma generate
npx prisma db pushnpm run dev- Update schema in
prisma/schema.prisma - Generate client →
npx prisma generate - Push DB changes →
npx prisma db push - Create UI components as needed
- Build API routes
- Connect frontend
- Test thoroughly
- Engagement → DAU, streaks, session time
- Time-Saving → Test generation <5s
- Learning Progress → Score improvement
- Accuracy → >90% CLAT pattern match
- Scalability → Thousands of learners concurrently
- Adoption & Retention → Registered vs active users
- Reduce dependency on costly offline coaching
- Empower aspirants with AI-curated adaptive prep
- Build a peer-learning ecosystem (rooms, leaderboards)
- Offer scalable, affordable, and exam-accurate preparation