Public UFO documents are complex, unstructured, and full of hidden relationships. This session uses them as a concrete case study to show how unstructured content can be transformed into a knowledge graph and connected to an AI agent for grounded, explainable answers.

We will move from raw documents to extracted entities, relationships, provenance, and uncertainty handling. Using Document Intelligence, we generate and inspect a graph model, explore it visually, query it, and finally connect it to an an Agent for natural-language interaction.