Glossary
/
AI Agents

RAG (Retrieval-Augmented Generation)

A method where an AI looks up documents before answering.

Definition

RAG is a method where an AI model looks up relevant documents, then uses them to write its answer. It works well for general questions and basic product discovery. It struggles with transactional tasks like orders, returns and complaints, which need structured, up-to-date data.

Key points

<ul><li>RAG finds text that looks similar to the question, so it can return the wrong SKU, an old price or an expired offer.</li><li>Many low-cost chatbots are mostly RAG over a dump of documents.</li><li>Accurate commerce agents combine retrieval with an ontology and a knowledge graph.</li></ul>