What is a Vector Database?
A vector database is a specialized database optimized for storing, indexing, and searching vectors — multi-dimensional numerical representations of text, images, or audio. Examples: Qdrant, Pinecone, Weaviate, Milvus, pgvector.
How do vectors work?
An embedding model converts text into a vector (e.g., 1,536 numbers). Semantically similar texts have "close" vectors in multi-dimensional space. The query "office in Warsaw" will be close to "business premises capital" despite different words — because their meaning is similar.
Role in RAG
Vector databases are the foundation of RAG architecture: documents are split into chunks, converted to vectors, and indexed. When querying, the system finds the most relevant fragments and provides them to the model as context.