Toward Knowledge-Centric Architectures for AI Agents

Today’s GenAI systems can synthesize and reason with impressive accuracy — up until you try to apply them to higher-stakes business problems. These introduce more challenges: high bars for accuracy and explainability, the need for just-in-time knowledge and decision context, and the ability to gate what data is used based on what is relevant and appropriate for a given purpose.
At the root of all this lies the question: how does your AI system gain access to your knowledge and context? In this talk, Neo4j’s CTO will talk about how agentic AI knowledge layers are being used to enable high-impact AI applications at companies like Uber, Adobe, Walmart, Abbvie, and many more. You will hear what the AI stack of the future looks like based on today’s trends, the need for an AI knowledge layer to work alongside your agents, and the necessary characteristics of this knowledge layer to support the next generation of agentic enterprise applications.