Tagged embeddings · show all articles
Retrieval is sold as a cheap fix for a model that does not know your data. Embedding compute, vector storage, ANN queries, reranking, and index freshness are five separate meters running underneath it, and none of them appear in the price of the model call.
Retrieval-augmented generation fails quietly. Ten distinct, documented mechanisms, from a stale index to a fabricated citation, account for nearly every expensive production RAG incident, and each has its own detection and its own fix.
An embedding is a filing decision, not a fact about meaning. Its geometry guarantees far less than the systems built on top of it assume, and every surprising retrieval failure follows from that gap.