Knowledge Sharing Portal

Full-stack learning portal with a React/Vite frontend and a Java Spring Boot backend. The app supports role-based access, subjects/classes, syllabus upload, Ollama-assisted content generation, subject Q&A, LLM dialogs, understanding analytics, and lightweight student peer threads.

Stack

• Frontend: React 19, TypeScript, Vite, Tailwind CSS
• Backend: Java 21, Spring Boot 3, Spring Security, JDBC
• Database: PostgreSQL with Flyway migrations
• LLM: Ollama-compatible local model API
• Auth: JWT bearer tokens
• File storage: local disk for v1

Prerequisites

• Node.js and npm
• Java 21
• Maven 3.9+
• PostgreSQL 15+
• Ollama running locally if you want live LLM generation

Backend Setup

Create a PostgreSQL database:

createdb knowledge_portal

Run the backend:

cd backend
mvn spring-boot:run

Default backend URL:

http://localhost:8080/api

Useful environment variables:

DB_URL=jdbc:postgresql://localhost:5432/knowledge_portal
DB_USERNAME=postgres
DB_PASSWORD=postgres
JWT_SECRET=change-this-development-secret-at-least-32-characters
UPLOAD_DIR=uploads
OLLAMA_URL=http://localhost:11434
OLLAMA_MODEL=llama3.1
CORS_ORIGINS=http://localhost:5173

Flyway creates the schema automatically on backend startup. Demo users are seeded when the database is empty:

Student: alex@student.com / student123
Teacher: sarah@quickquery.com / teacher123
Admin: admin@quickquery.com / admin123

Ollama Setup

Install/pull a model:

ollama pull llama3.1
ollama serve

If Ollama is unavailable, syllabus processing still stores fallback generated content with ready_for_review status so the workflow remains testable.

Frontend Setup

Install dependencies:

cd frontend
npm install

Create .env.development or update the existing file:

VITE_API_URL=http://localhost:8080/api
VITE_APP_ENV=development
VITE_ENABLE_DEBUG=true

Run the frontend:

cd frontend
npm run dev

Default frontend URL:

http://localhost:5173

Implemented Backend APIs

• GET /api/health
• POST /api/auth/register
• POST /api/auth/login
• GET /api/me
• GET /api/classes
• POST /api/classes
• GET /api/subjects
• POST /api/subjects
• POST /api/subjects/{subjectId}/enrollments
• POST /api/subjects/{subjectId}/syllabi
• GET /api/subjects/{subjectId}/syllabi
• POST /api/subjects/{subjectId}/materials
• GET /api/subjects/{subjectId}/materials
• GET /api/subjects/{subjectId}/generated-content
• PATCH /api/generated-content/{contentId}
• POST /api/syllabi/{syllabusId}/regenerate
• GET /api/subjects/{subjectId}/questions
• POST /api/subjects/{subjectId}/questions
• POST /api/questions/{questionId}/replies
• GET /api/subjects/{subjectId}/conversations
• POST /api/subjects/{subjectId}/conversations
• POST /api/conversations/{conversationId}/messages
• GET /api/peer-threads
• POST /api/peer-threads
• POST /api/peer-threads/{threadId}/replies
• POST /api/subjects/{subjectId}/understanding
• GET /api/subjects/{subjectId}/understanding/me
• GET /api/subjects/{subjectId}/analytics/understanding

Verification

Frontend type check:

cd frontend
npx tsc -b

Frontend production build:

cd frontend
npm run build

Backend tests/build:

cd backend
mvn test

Current Limitations

• PDF and image text extraction are placeholders. Add Apache PDFBox and OCR when richer document ingestion is needed.
• LLM processing is synchronous for v1. Move syllabus processing to a background job queue before production use.
• Local disk file storage is suitable for development only.
• Peer threads are intentionally minimal and do not include advanced moderation or real-time updates.
• Generated content requires teacher approval before students see it as approved material.