LLMs are trained on vast amounts of text to learn the statistical patterns of language — which words and ideas tend to follow others. This training gives them the ability to understand a wide range of phrasing for the same underlying request, and to generate responses that read naturally. In ecommerce, LLMs are what make it possible for a shopper to type "something for my sister's wedding, nothing too flashy" and have the system correctly interpret occasion (wedding), formality (moderate), and constraint (not flashy) — rather than requiring the shopper to fill out a structured form. LLMs alone are not sufficient for a good shopping experience. An LLM understands language; it does not know a brand's catalog, sizing, or inventory. Production AI stylists combine an LLM's language understanding with a structured reasoning layer connected to real catalog and taste data — which is where systems like Elara's Style Graph and outfit assembly logic come in.