You don't just feed documents in, you need to build a dataset representative of how you want to interact with it. So likely using gpt-4 or something to create:

You don't just feed documents in, you need to build a dataset representative of how you want to interact with it. So likely using gpt-4 or something to create: a chunk of a document, a question that can be answered by that chunk and a good answer. (Or something)