Vector Database

A vector database is a specialized database type that stores and indexes data (text, images, audio) as mathematical vectors — i.e., multi-dimensional arrays of numbers (embeddings). This structure enables semantic search, where queries are based not on keyword matching but on meaning-based similarity (distance). Vector databases are indispensable for RAG systems and modern recommendation engines, as they enable fast and contextual retrieval of vast amounts of unstructured data, providing long-term memory for AI applications.