Pristine knowledge content is essential for evolving from simple conversational chatbots to fully agentic AI agents that act on customers’ behalf.
Metrigy’s latest customer experience (CX) research shows that the vast majority of 393 participating companies have either already implemented or are piloting the use of basic, structured conversational AI, generative AI, and agentic AI for customer interactions involving knowledge content.
Presumably, the “already implemented” projects fall within the realm of structured, conversational AI characteristic of chatbots, and perhaps to a lesser degree generative AI. Unless companies are absolutely certain that pointing AI agents to knowledge content won’t lead to the sharing of bad information, then they shouldn’t be entertaining a production deployment. (And by “bad” information, I mean content that’s inaccurate, invalid, irrelevant, or otherwise potentially problematic.) It’s a tall order, and kudos to the many companies that have prioritized a knowledge clean-up in support of planned generative and/or agentic AI use cases.
Continue reading at nojitter.com.
