I-GYAN — Tech Stack & System Structure
1. Frontend (Client-Side)
Goal: Deliver a fast, modern, mobile-first user experience for students, mentors,
and HRs.
Stack:
Tech Purpose
React.js Component-based UI architecture
Tailwind CSS Utility-first styling framework
React Router Page-based routing (SPA)
Axios / Fetch API Communication with backend
React Context / Zustand Global state management
Framer Motion Smooth animations
React Hook Form + Zod Form handling + validation
Firebase Auth SDK (Optional) OAuth for login/signup
PWA (Optional) Offline support, mobile-friendliness
How it works:
• The homepage loads a responsive UI with sections like Hero, Courses,
Projects, Mentors, etc.
• User interacts with AI Chat Bar (goal input)
• AI responds with roadmap → user directed to relevant course/project
• Authenticated users can submit projects, talk to mentors, or receive
feedback
2. Backend (Server-Side)
Goal: Handle business logic, API responses, AI routing, and DB interaction
Stack:
Tech Purpose
Node.js + Express RESTful API framework
Prisma ORM / Mongoose Database communication
Zod / Yup API input validation
JWT / Firebase Auth Secure user authentication
LangChain / RAG AI orchestration (optional)
Socket.IO Real-time chat & notifications
How it works:
• API routes like /api/users, /api/roadmap, /api/projects, /api/messages
• Auth middleware checks user tokens
• AI-related routes communicate with OpenAI API
• Admin-specific routes secured with role-based access
3. Database
Goal: Persist all user data, projects, feedback, skills, and relationships
Stack:
Tech Purpose
PostgreSQL Relational database (preferred for structured queries)
MongoDB (optional) Flexible NoSQL for content-heavy systems
Key Tables / Collections:
• Users (students, mentors, HRs)
• Skills & Tracks
• Projects (submitted + templates)
• AI Chats / Roadmaps
• Mentorship Feedback
• Badges / Certificates
• Messages
4. AI/ML Engine
Goal: Drive personalized roadmaps, career chat, resume review, interview bot
Stack:
Tool Purpose
OpenAI GPT-4 / Gemini Pro Natural language generation
LangChain Multi-agent prompt routing
Pinecone / ChromaDB (Optional) Vector store for retrieval
Custom Prompts Based on student goal + skill level
How it works:
• User types: “I want to become a UI/UX Designer”
• System generates:
o Course path
o Tools to learn
o Weekly goal sheet
o Micro-projects
• Responses cached for faster delivery
5. Storage & Media
Goal: Store user-uploaded project files, images, and profile assets
Stack:
Tech Purpose
Firebase Storage Easy media handling
AWS S3 Enterprise-grade file storage
6. Real-Time & Messaging
Goal: Enable student–mentor chat, HR–student feedback, notifications
Stack:
Tech Purpose
Socket.io WebSockets for bi-directional messaging
Firebase Realtime DB Alternative for small-scale real-time logic
Push Notifications Web push for updates, reminders
7. Authentication & Roles
Goal: Secure access for different roles (student, mentor, HR, admin)
Stack:
Tech Purpose
Firebase Auth / JWT Secure login/signup
Role-based Access Middleware Protect sensitive routes
Google OAuth / LinkedIn OAuth Easy signup for HRs/mentors
8. Admin Panel & Analytics
Goal: Track platform growth, usage, and manage users/content
Stack:
Tool Purpose
React Admin / custom dashboard Admin UI to manage system
Mixpanel / PostHog Track user behavior & actions
LogRocket / Sentry Error tracking and UX analysis
9. DevOps / Deployment
Goal: Automate build, deploy, and scale securely
Stack:
Tech Purpose
Vercel / Railway Frontend & backend hosting
GitHub Actions CI/CD pipeline
Docker (optional) Containerization for scalable hosting
Cloudflare DNS, security, and CDN
10. Flow Summary
User Visits Site → Inputs Goal → AI
Generates Roadmap →
User Learns or Submits Project → Mentor
Reviews → HR Can View →
User Builds Public Profile → Gets Hired or
Recognized
