AI That Executes the Language of the Enterprise: An Ontology-Based Execution Layer
Session Overview
Generative AI has become fast and remarkably capable. But the moment you bring it onto the enterprise floor, it runs into a wall. The model speaks fluently β yet it can't actually get the work done in your company's own language. Whether it's product code, transactional relationships, or financial rules, it produces plausible-sounding answers while stalling at the point of real execution.
This talk addresses where true competitive advantage accumulates in an era when models are becoming commoditized. Through an ontology that structures enterprise knowledge, an execution layer that performs work on top of it, and a decision lineage that traces the rationale behind every action, it shows β with real-world examples β how AI agents can move beyond "talking" to becoming the actor that "gets things done."
Speaker
Seunghyun Lee is an AI startup CEO. He earned a master's in computer science at KAIST and led Bixby Vision at Samsung Electronics, driving AI vision technology. He went on to found the AI-Native service company Enhans, helping enterprises transform into Agentic Enterprises where AI agents work and make decisions autonomously. Through its AgentOS platform, Enhans is leading the AI transformation of major Korean companies including Samsung, LG, and SK, and is an active member of the Microsoft AI Startup Alliance.