harnyGlossary
Vector memory
Vector memory stores information as numeric embeddings and finds it by similarity. An AI agent turns a piece of text into a vector, then searches for stored vectors that sit close to the query.
What vector memory is
Vector memory is a store of text turned into embeddings, which are lists of numbers that capture meaning. A query is embedded the same way, and the closest stored items are returned.
Agents use it to recall past conversations, notes and documents without reading everything again.
Why it matters and why it is hard
- It finds related text even when the wording differs.
- Similar facts look alike to a search, so an old value can rank above a newer one.
- Similarity alone cannot say which of two facts is current.
How it works in Harny
- Memory is dated: a changed value with a newer timestamp becomes the current value, and the earlier value stays answerable by date.
- You recall with
POST /v1/memory/recall, and you can sort by relevance or chronologically.
Does vector memory know which fact is newest?
Not by similarity alone. Your app needs dates or replacement to decide which value is current.
Keep reading
Go on. Get Harny.
The first 500 monthly-active end users are free for good. Create an app key and make your first call in minutes.