Welcome to my Streamlit app designed for invoice extraction utilizing OpenAI's API. This versatile tool offers two distinct output models for enhanced flexibility and accuracy in extracting invoice data.
Acknowledging the widespread demand for data extraction solutions utilizing LLM, and inspired by Langchain's introduction of the with_structured_output
method for enhanced model calls, I centered my efforts on tackling the common use case of invoice extraction, culminating in the development of this bot.
- Supports extraction of data from
PDF
files - Utilizes
function calling
with LLM to enhance accuracy - Offers two output options:
JSON
andCSV
Extracter-bot-demo.mp4
In upcoming iterations, I aim to incorporate the ability to extract data from Excel and Word files, expanding the application's versatility and utility.
Clone the repository
git clone [email protected]:Ja-yy/Invoice-excreter-bot-llm.git
Create a virtual environment using Pipenv:
pipenv install
Start Streamlit server:
streamlit run main.py
Now, go to localhost:8501
Enjoy the app :)