NODES AI 2026 - Temporal Substrate Architecture: Building Persistent Identity for Autonomous Agents

In this talk, Conor O'Shea presents "Temporal Substrate Architecture: Building Persistent Identity for Autonomous AI Agents" at NODES AI. The session explores the concept of AI agents developing a persistent identity through learning, contrasting traditional memory retrieval methods with a new approach that fosters a deeper sense of self. By employing Neo4j, O'Shea showcases a production architecture that enables AI agents to maintain continuity and build upon past experiences, illustrated through an autonomous network security analyst navigating enterprise firewall ontologies. This innovative use of graph-native memory shifts agent behavior from mere response generation to a more nuanced understanding shaped by accumulated experiences.