VectorDB overview
What vector search is in plain words, when to use Crusoe VectorDB, and how it compares to Vertex AI Vector Search, OpenSearch, and Azure AI Search.
VectorDB quickstart
Create a vector index, insert points, and run your first similarity search in about five minutes.
Indexes and points
The VectorDB data model — indexes, points, vectors, payloads, ids, dimensions, distance metrics, and what can never change.
Search
Query patterns for VectorDB — nearest-neighbor search, payload filters, top-k, scores, browsing with scroll, and deleting points.
Use with agents
How the agent memory bank uses vector search — mem_ collections, memorize, and the search_memory tool — and when to use VectorDB directly instead.