A Streamlit application that uses Google's Gemini AI to extract structured data from documents based on user-defined schemas.
-
Dynamic schema builder interface
-
Support for multiple field types:
- String
- Integer
- Float
- Boolean
- List variations (String, Integer, Float, Boolean)
-
File upload capability
-
Automated data extraction using Gemini AI
-
Real-time JSON output
- Built with Streamlit for the web interface
- Uses Pydantic for data validation and schema generation
- Integrates with Google's Gemini AI for document processing
- Supports dynamic schema creation based on user input
- Clone the repository
- Install the required dependencies
- Add your Gemini API key in the .env
- Run the Streamlit app -
streamlit run app.py