Singularity Academy

A full-stack AI-powered education platform combining modern web technologies with advanced AI capabilities for personalized learning experiences.

System Architecture

graph TD
    A[Client] --> B[Caddy Reverse Proxy]
    B --> C[Frontend: React/TS]
    B --> D[Backend: Go/Gin]
    B --> E[AI Engine: Python/FastAPI]
    D <--> F[(MySQL Database)]
    E --> G[LangChain Agents]
    G --> H[OpenAI Integration]
    E --> I[Manim Visualization]
    E --> J[Document Processing]

Loading

Technology Stack

Core Components

• Reverse Proxy: Caddy 2.6+ with automatic HTTPS

• Frontend:

  • React 18+ with TypeScript 5+
  • State Management: Zustand 4+
  • Styling: Tailwind CSS 3.3+ with Headless UI
  • Routing: React Router 6.14+

• Backend (Go):

  • Gin Web Framework 1.9+
  • GORM 2.0+ ORM
  • JWT Authentication
  • WebSocket Integration

• AI Engine (Python):

  • FastAPI 0.95+ with WebSockets
  • LangChain 0.0.340+ for AI agent orchestration
  • Manim Community 0.17+ for mathematical visualizations
  • Document Processing: PyMuPDF, python-docx, python-pptx

AI Capabilities

• Dean AI Agent:

  • Dynamic study plan generation
  • Multi-stage conversational interface
  • Document-aware curriculum planning
  • Manim animation script generation

• Document Processing:

  • Supported formats: PDF, DOCX, PPTX, CSV
  • Text extraction and semantic analysis
  • Content chunking for AI processing

Database

• MySQL 8.0+ with InnoDB engine
• Database schema versioning with migrations
• Connection pooling with Go-MySQL-Driver

DevOps

• Containerization: Docker 20.10+
• CI/CD: GitHub Actions
• Monitoring: Prometheus + Grafana
• Logging: Loki + Promtail

Key Features

AI-Powered Learning Flow

1. Document Upload & Processing:

   • Secure file validation and storage
   • Multi-format text extraction
   • Content chunking for AI analysis

2. Dynamic Study Plan Generation:

   async def generate_study_plan(self) -> Dict:
       # Combines document content with pedagogical strategies
       # Generates interactive learning modules with:
       # - Manim animation scripts
       # - Conversational prompts
       # - Assessment checkpoints

3. Real-Time AI Interaction:

   • WebSocket-based communication
   • Context-aware conversation management
   • Progressive content delivery

Security Features

• JWT-based authentication
• File validation:

  def fetch_uploaded_files(self) -> str:
      # Security checks:
      # - User-specific file naming
      # - Path validation
      # - File existence verification

• Rate limiting and request validation
• HTTPS-only communication

Getting Started

Prerequisites

• Node.js 18.12+ (LTS)
• Go 1.20+
• Python 3.11+
• MySQL 8.0+
• Redis 7.0+ (for caching)

Installation

AI Engine Setup

cd backend/ai_engine
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

# Configure environment
cp config/config.example.json config/config.json
# Add your OpenAI API key

Go Backend Setup

cd backend
go mod download

# Configure environment
cp .env.example .env
# Update database credentials

# Run migrations
go run cmd/migrate/main.go

Running the System

# Start AI Engine
uvicorn main:app --reload --port 8000
uvicorn main:app --reload --port 8000

# Start Go Backend
go run .

# Start Frontend
cd frontend
npm install --force
npm start

# Caddy Proxy Server: At root
caddy start

API Documentation

AI Engine Endpoints

• WS /ai/dean_ai/{token}?material={filename}
  • Real-time study plan generation
  • Payload format:

  {
    "manim_script": "Python code for visualization",
    "message": "Natural language explanation",
    "notes": "Key learning points"
  }

Core Endpoints

• Authentication: JWT-based
• File Management:
  • POST /api/materials
  • GET /api/materials/{id}
• Course Management:
  • POST /api/courses
  • GET /api/courses/{id}/study-plan

Development Guidelines

Branch Strategy

• main: Production-ready code
• develop: Integration branch
• Feature branches: feature/[description]

Testing

• Go: Native testing package
• Python: pytest with 90%+ coverage
• Frontend: Jest + React Testing Library

License

No current license, since it's not an open source project.

Roadmap

Q1-2 2025

• Multi-modal AI integration (vision/audio)
• Real-time collaboration features
• Adaptive assessment engine

Q3-4 2025

• Mobile app integration
• AI teaching assistant marketplace
• Blockchain credential verification

1. System architecture visualization
2. Detailed technology specifications
3. Key implementation details from the code
4. Clear setup instructions
5. API documentation
6. Development guidelines
7. Future roadmap