Vector Database
Definition pending
We've cataloged "Vector Database" but haven't written a full definition yet. Definitions are hand-curated rather than auto-generated, so it takes time to cover every term.
Want this one prioritized? Email us and we'll bump it.
Practical example
A vector database stores embeddings (arrays of floats) and lets you search by similarity. For LLM apps, it's the memory layer: store document embeddings, retrieve relevant ones at query time, inject into prompt. Choices: Pinecone (managed, $$), Weaviate (self-host, open-source), Qdrant (Rust, fast), Chroma (lightweight, embedded), pgvector (Postgres extension, simplest if you already run Postgres).
Workflow example
Pick your vector DB: if you're under 100K documents, use pgvector — no new service to manage. If 100K–10M documents with <50ms P99 latency requirement, use Qdrant or Weaviate. If you need managed + scale, Pinecone. Operational checklist: (1) set up embedding pipeline (batch + real-time), (2) choose embedding model (768-dim is standard), (3) define index type (HNSW for most cases), (4) monitor index size, QPS, and recall@k. Re-index when recall drops.