Notifications
Clear all
Topic starter
18/07/2026 4:48 am
I'm starting a project to build a Q&A chatbot over a collection of roughly 10,000 PDFs. The goal is to let users ask questions and get answers based on the content.
I plan to use LlamaIndex as the core framework. For those who have done this at a similar scale, what specific tools and components worked best for you? I'm especially unsure about the best PDF parser, embedding model, and vector database to pair with it for this volume of data. Any guidance on the overall architecture would be a huge help.