Question: Best RAG strategy for 12 heavy Textbooks?

I have a client project using n8n and I need some advice on the RAG setup.

The client has 12 large textbooks (PDFs). I need to build an automation where the AI generates course lessons, but it must use only specific books or chapters as the source material (not the whole library at once).

I know that you will teach us about this in the upcoming course, but I just want to quickly understand it because I need it right now.

My questions:

What is the best strategy to "chunk" or split these large files? Should I split them by chapter before uploading?
How do I set up the metadata so the AI knows which book is which?
Are there any specific tutorials or resources you recommend for building a "Textbook RAG" system in n8n?

Thanks in advance!

6 comments