What is large language model (LLM)?
A large language model is an AI system trained on very large amounts of text so that it can read and write natural language. It produces a reply by predicting the words most likely to come next.
Why it matters
Large language models are what allow current AI agents to understand loosely worded questions and write replies that read naturally. They can summarise a long conversation, draft an email or classify a ticket without being programmed for each task. Knowing how they work also helps a team understand their limits, including the risk of confident but wrong answers.
How it works
During training, the model sees a vast amount of text and learns statistical patterns in how words follow one another. When it receives a prompt, it generates a reply one small piece at a time, each choice based on everything that came before. It has no built-in knowledge of a particular organisation, so support systems supply relevant documents and instructions alongside the customer’s message.
Example
A support agent pastes a long email thread into an assistant and asks for a three-line summary. The model reads the thread and writes one. If the agent then asks what the company’s refund period is, the model can only answer correctly if that policy has been supplied to it. Otherwise it may guess, which is why answers are grounded in approved content.
