Retrieval
Embedding
A numeric representation of meaning, letting a computer find text that is similar in sense rather than in wording.
An embedding turns a piece of text into a list of numbers positioned so that similar meanings sit close together. "Cancel my order" and "I'd like a refund" end up near each other even though they share almost no words.
That is what makes semantic search possible, and it is the foundation of RAG.
Embeddings are cheap compared with generating text, which is why well-designed systems use them to narrow the field before involving an expensive model. They are also weaker at exact matches — an embedding will not reliably find part number XR-4471. That is why serious systems combine them with keyword search. See hybrid search.