An embedding maps a piece of text to a fixed-length list of numbers (a vector): content with similar meaning ends up close together in that space. It lets machines measure how alike two passages are by distance, instead of only comparing literal words.
In Enclave, embeddings underpin long-term memory and knowledge retrieval: conversations, events, and documents are encoded as vectors, so the system can recall the most relevant memories by meaning when needed, rather than relying on exact keyword matches.
Try it yourself in Enclave
Open it in your browser — no credit card, no install.