Infrastructure

Vector Database

A specialized database designed to store and search data represented as mathematical vectors (embeddings).

Detailed Explanation

A Vector Database is purpose-built to handle high-dimensional vector embeddings generated by AI models. Unlike traditional relational databases that search for exact keyword matches, vector databases enable 'semantic search'. They can find data that is conceptually similar to a query by measuring the mathematical distance between vectors. This technology is the backbone of RAG systems.

Related Terms

Embeddings

Mathematical representations of data (like words or images) in a high-dimensional space.

RAG (Retrieval-Augmented Generation)

A technique that connects an LLM to external knowledge bases to improve accuracy and reduce hallucinations.

Semantic Search

A search technique that looks for the meaning and context of a query rather than exact keyword matches.

Back to Glossary