5-Minute Learning App

Tired of LLMs hallucinating or over-explaining simple concepts—like using 1500 words to justify why 8.11 is greater than 8.9? Meet your solution: a powerful web app that delivers concise, accurate overviews of any topic to understand in just 5 minutes, powered by intelligent multi-LLM summarization and comparison.

🌟 Core Features

1. 5-Minute Learning Experience

Ask any question and get a comprehensive, digestible summary in about 5 minutes. Perfect for quick learning sessions on topics like:

• "The difference between PPO and GRPO"
• "The impact of the appreciation of the US dollar"
• "The history of Nike vs Adidas"

2. Multi-LLM Summarization

• Expert Consultation: Queries multiple LLM providers (ChatGPT, Claude, Gemini) simultaneously
• Intelligent Summarization: Uses local Ollama models to compare and synthesize responses and save cost
• Consensus Highlighting: Identifies consistent information across sources and points out differences
• Source Transparency: Shows which models contributed to each answer

3. AI-Powered Development with Cline

This project showcases modern AI-assisted development practices:

• Built using Claude's Cline for intelligent code generation
• Comprehensive test coverage with both unit and integration tests
• Automated database management and visualization tools

🎬 Demo

🚀 Installation

Prerequisites

• Python 3.8+
• Node.js 16+
• PostgreSQL 12+
• Ollama (for local LLM)

1. Clone the Repository

git clone https://github.com/yourusername/5mins-learning-app.git
cd 5mins-learning-app

2. Backend Setup

Install Python Dependencies

# Create virtual environment (recommended)
python -m venv venv
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

Database Setup

# Create PostgreSQL database
createdb 5mins_learning

# Set up database schema and test data
./db/reset_db.sh

Environment Configuration

# Copy environment template
cp .env.example .env.local

# Edit .env.local with your settings

Required environment variables:

# Database
DATABASE_URL=postgresql+asyncpg://username@localhost:5432/5mins_learning

# Optional API Keys (leave empty to disable specific providers)
OPENAI_API_KEY=your_openai_key_here
ANTHROPIC_API_KEY=your_anthropic_key_here
GEMINI_API_KEY=your_gemini_key_here

Install and Configure Ollama

# Install Ollama from https://ollama.ai/
brew install ollama

# Pull required models
ollama pull gemma3:4b
ollama pull gemma3:12b

# Verify Ollama is running
curl http://localhost:11434/api/tags

3. Frontend Setup

cd frontend
npm install

🎯 Usage

Starting the Application

1. Start Backend Server

# From project root
uvicorn app.main:app --reload

Backend runs on: http://localhost:8000 API documentation: http://localhost:8000/docs

2. Start Frontend Development Server

# From frontend directory
cd frontend
npm start

Frontend runs on: http://localhost:3000

Using the Application

Home Page

• Enter your learning question in the central input box
• Select expert LLM providers (ChatGPT, Claude, Gemini)
• Choose your local summarization model (gemma3:4b or gemma3:12b)
• Click submit to start your 5-minute learning session

Conversation Page

• View real-time responses as they stream in
• See which models contributed to each response
• Toggle visibility of the AI's thinking process
• Ask follow-up questions with model reselection
• Use "Retry" to regenerate responses with different models

History Page

• Browse all previous conversations
• Quick access to continue past discussions
• Delete conversations you no longer need

🛠️ Development

Database Tools

# Reset entire database
./db/reset_db.sh

# View database contents
python -m db.visualize_db

# View specific conversation
python -m db.visualize_db --conversation [ID]

# View specific message
python -m db.visualize_db --message [ID]

Testing

# Run unit tests
pytest app/tests/mock_tests/

# Run integration tests (requires running server)
python -m app.tests.real_tests.test_live_server

Key API Endpoints

• POST /api/conversations - Create new conversation
• POST /api/conversations/{id}/first-reply - Generate initial response
• POST /api/conversations/{id}/reply - Continue conversation
• GET /api/conversations/{id}/messages - Load conversation history

🔧 Configuration

LLM Provider Configuration

The app supports flexible LLM provider configuration:

• Multi-provider queries: Select any combination of ChatGPT, Claude, and Gemini
• Local summarization: Choose between gemma3:4b (faster) or gemma3:12b (more capable)
• Graceful degradation: App works with any available providers

Model Options

• OpenAI: gpt-4o-mini (default)
• Anthropic: claude-3-haiku-20240307 (default)
• Google: gemini-2.0-flash (default)
• Ollama: gemma3:4b, gemma3:12b, deepseek-r1:8b

🤝 Contributing

Welcome any kind of contribution.

License

Made with ❤️ for curious minds who value efficient learning and AI coding