Similarity is a product choice
Vector search feels objective because it returns scores. But those scores come from a model trained to notice some kinds of similarity more than others. Your corpus, chunking, and user language decide whether that is helpful.
The useful signal is usually already there.
Hit the button and watch the hidden AI surface area light up.
Look at the neighbors
Do not tune embeddings from aggregate scores alone. Inspect nearest neighbors for real queries, label misses, and decide which mistakes matter to the product.
A simple neighbor dump can explain a lot of strange answers.
Pair vectors with old-school filters
Metadata filters, keyword search, recency rules, and reranking can make vector search less mysterious. The best retrieval systems are usually hybrids.
Flip the switches and watch the story turn into a tiny operating model.
Numbers still need judgment
Embeddings are powerful because they compress meaning. They are risky for the same reason. Keep the compression visible.