In this presentation, I explain how to build a document processing pipeline to create structured data for usage in LLMs.
This version of the talk was given at IBM TechXchange, in October 2025.