llama-hub
has a tool for. But even if you didn't use llama-hub tools for it so as long as you parsed it to extract the text, images, and other objects from the pdf and could store that info in a vector database for retrieval you can do a lot with llama index.