“Looks good to me!
I think it's cool!
- The dev of this tool

How each process works?
PDF
The PDF files uploaded get converted to an image using `react-pdf` and is read by Optical Character Recognition via `tesseract-js`.
DOCX
These files are run through `mammoth` which is a dependency extracts the text from the file.
PPTX
PPTX files are made from XML files so the only process necessary is using `JSZIP` to unzip and loop through the slides and cut out formatting elements.
Summarization
The AI used is hosted in Hugging Face and switches models based on the length of the file. Uses `sshleifer/distilbart-cnn-12-6` and `pszemraj/led-large-book-summary`