New: Cookbooks and AI ExplanationsStep-by-Step recipes to solve problems connected to Roadmaps and Cheat Sheets. Need more details? Use AI buttons for structured and simple explanations with concrete examples throughout the whole platform.Take a look
I know I wrote something down somewhere, but I can't remember the exact words, and keyword search finds nothing.
What you'll have at the end
A script that returns your three most relevant notes for a plain-language query, ranked by meaning.
You need
A working embedding call that already returns a vector for a piece of text, and a folder of your own short, plain-text notes with real entries to search over.
Not covered
Keeping the saved index in sync automatically as notes are added or edited, and embedding a note too long for one call; this only covers a first working search over a fixed set of short notes.
Leans on
Verify your first embedding works
go there first if you don't yet have a working embedding call to check before pointing it at real notes
Choose cosine similarity or dot product for your vectors
go there if your embedding model isn't confirmed normalized, since this recipe assumes cosine similarity is the right default
Speed up nearest neighbor search with a FAISS index
go there once your own notes grow past what scoring every single one on each search comfortably handles
Checked 15 Aug 2026
Part of the Embeddings cookbook