Terminology data is one of the most powerful guardrails to keep AI-generated content on-brand and factually sound, just as it has been guiding human-generated content quality for decades. The game-changing opportunity for terminology data is to inject not only corporate brand language, but also contextual knowledge into AI output, increasing the value terminology management brings to the entire company.
We call this approach Terminology-Augmented Generation, based on the more commonly known Retrieval-Augmented Generation. However, terminology resources have often not been maintained optimally or have simply not been optimized for large language models, limiting their potential.
Join terminology veterans Barbara Inge Karsch and Kara Warburton as they tackle the uncomfortable truth about legacy terminology assets and what it takes to get them into AI-ready shape. Together with Klaus Fleischmann as a moderator, they will explore proven methods to bridge the quality gap, investigate how AI can also help in terminology work, and share hard-won, real-life experiences from the terminology trenches.
