Everyone is rushing to deploy AI into contact centres and Microsoft Teams workflows, but very few people understand what these systems are actually doing under the hood. There is a common misconception that AI acts as a hyper-fast search engine, looking up facts and citing them. It isn’t.

In this 25-minute, non-technical dive, James Cadman skips the buzzwords to explain the reality of Large Language Models (LLMs). In plain English, James will unpack how AI actually “thinks” by predicting the next most probable word – acting as the world’s most advanced autocomplete rather than a traditional database.

By understanding the underlying mechanics of how these models ingest context, analyse semantic meaning, and generate responses, you will finally understand why AI hallucinates, how it interprets a customer’s intent, and what it takes to ground it in your company’s reality.

Using practical contact centre examples, you’ll walk away with a clear mental model for how these systems actually process language. This foundational knowledge will empower you to separate hype from reality, challenge vendors more effectively, and design AI use cases that genuinely improve both CX and the human agent experience. This session is designed for business and technical audiences alike: no complex math or academic jargon, just transparent insight into the machine’s logic.