I scanned my entire home library and have all the books in pdf format. Most of them can’t be found on the internet. In addition to books on history, art, history and fiction, I also have books on medicine, mathematics, physics, chemistry, etc.
I need to scan my PDF books, to make a database, then starting ChatGPT for getting answers. I have to use API, to crawl all books.
Right now, to complete this task, and to get good results, I’m using Mathpix Snip (mathpix.com) , to scann each PDF file with OCR and LATEX, then save each file in txt. Mathpix Snip is the best tool than can convert mathematics, physics and chemistry books into latex, and save as txt files. After that, I connect txt database with API from OpenAI. Simple, but…
The problem: it takes a lot of time !
So, I need all these procedures to be done much faster, at once: scanning pdfs, converting them to txt files, to make a unique database of 500 books, so that I can get start using ChatGPT focus only to my database.
So, how can I scann, convert with OCR+LATEX, using API from OpenAi, and crawling all books to make a database, then start ChatGPT prompt for answers?