NODES AI 2026 - How to Build the Next Wave of Graph-Based Agentic AI

This talk features Charles Borderie at NODES AI, focusing on the future of Graph-Based Agentic AI. It addresses the limitations of current generative AI agents, such as hallucinations and memory issues, emphasizing the importance of structured reasoning over mere large language models. The speaker explores the architecture needed for reliable AI, highlighting how Knowledge Graphs serve as the foundational structure for precision reasoning. By analyzing Lettria’s Perseus model, this session provides insights into overcoming the "Text2Graph" challenge and achieving high validation rates, as well as strategies to reduce latency for real-time applications. Attendees will learn practical methods for combining Document Parsing, Ontology Enforcement, and GraphRAG to develop transparent and enterprise-ready agents.