[constantinum]

The tool doesn't use any LLMs for processing/parsing the data. It parses and converts into raw text.

The final output(raw text) of the parsing is then fed to LLMs for data extraction. e.g. Extracting data from insurance, banking, and invoice documents.