Vector Database
A specialised database that stores text or other content as numeric embeddings and supports similarity search.
In detail
A vector database stores embeddings, the high-dimensional numeric representations that AI models produce for text, images or audio. The database supports nearest-neighbour search, so given a query embedding it returns the most similar stored embeddings. This is the storage layer for retrieval-augmented generation and semantic search. Common options include Pinecone, Weaviate, Qdrant, Postgres with pgvector and Supabase's vector extension. Each makes different trade-offs on hosting, cost and query speed.
Why it matters for Australian business
For Australian businesses building an internal AI assistant, the vector database is where your indexed knowledge lives. Choice matters for data residency (Pinecone offers Australia, others do not), cost at scale, and operational complexity. Many small Australian deployments do well on Postgres with pgvector hosted on a Sydney AWS region, which keeps everything in one familiar database and inside Australian data sovereignty boundaries.