Hello!
I've been on a off these last months on a project to feed a bunch of my books to AI and let it figure out accurate bibliographical information about cited authors and books.
This is done with a fairly complex AI pipeline that extracts citations from the text of my books, it then tries to find it both on Wikipedia and Goodreads to resolve information like original publication dates, birth/death/nationality of the author and a lot more stuff!
It costs some money to run every single book since those API calls are a bit expensive so for now I ran the system on ~40 books I had on calibre!
Code is here if any coder wants to take a look! LINK
by thiago_lira