A database built to store embeddings, the numeric fingerprints of text or images, and to find the ones closest in meaning to a query.
A Vector Database stores embeddings, the long lists of numbers a machine learning model produces to capture the meaning of a passage of text or an image.
Unlike traditional relational databases that search for exact keyword matches, vector databases perform "similarity searches." They can rapidly find the vectors that are mathematically closest to a given query vector.
Retrieval-Augmented Generation (RAG) is built on this capability. When an enterprise Large Language Model needs to look up corporate documents to answer a question, it queries a vector database for the most semantically relevant information.
See also Machine Learning, Retrieval-Augmented Generation, Large Language Model, Data Mobilization, Data Lakehouse and Data Sovereignty.