What happens in the real agent?

• Extracts text with PyMuPDF (fast PDF parsing).
• Classifies with GPT (invoice, contract, cv, unknown).
• Renames as {type}__ai__{original}.pdf and routes to a folder.
• Writes a summary.csv of all actions.

git clone https://github.com/oaibuilds/O-Reader
cd O-Reader
python -m venv .venv && source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install -r requirements.txt
echo OPENAI_API_KEY=sk-xxx > .env
python main.py

Privacy: this demo does not upload files. Everything runs in your browser to show the UX idea.