Phoenix - LLM Flashcard Generator

Phoenix is a Streamlit app that generates editable flashcards from study material. It accepts pasted text or uploaded PDF, DOCX, and TXT files, then uses Gemini structured outputs with Pydantic validation to return a predictable flashcard format.

What It Does

• Upload .pdf, .docx, or .txt files, or paste raw text.
• Choose a subject to guide the flashcard style.
• Generate question-answer flashcards using Gemini.
• Validate the LLM response with Pydantic models.
• Review and edit flashcards in the UI.
• Export flashcards as CSV or JSON.

Tech Stack

• Streamlit
• Python
• Google GenAI SDK
• Pydantic
• PyPDF2
• python-docx
• Docker
• pytest

Project Structure

Phoenix/
├── Dockerfile
├── README.md
├── requirements.txt
├── src/
│   ├── app.py
│   ├── models/flashcard.py
│   ├── services/llm_service.py
│   └── utils/document_parser.py
└── tests/test_gemini_flashcards.py

Local Setup

git clone https://github.com/techwallahexplorer/Phoenix.git
cd Phoenix
cp .env.example .env
pip install -r requirements.txt
streamlit run src/app.py

Set your Gemini key in .env:

GEMINI_API_KEY=your_gemini_api_key_here

Docker

docker build -t phoenix-flashcards .
docker run -p 8501:8501 -e GEMINI_API_KEY=your_gemini_api_key_here phoenix-flashcards

Testing

pytest tests/ -v

The test suite mocks Gemini calls so CI does not depend on live API keys.

Notes

This project is a learning/product prototype. It demonstrates document parsing, LLM structured output handling, validation, and simple export workflows.