SQL vs NoSQL Databases: How to Choose
SQL vs NoSQL explained by data shape and access pattern, with concrete examples for relational, document, key-value, wide-column and graph databases.
Data & Analytics
What is Vector Database?
Quick answer
A database designed to store embeddings and quickly find the items most similar to a given query vector.
Vector Database, explained
Traditional databases are excellent at exact matches, such as finding a customer by ID. Vector databases specialize in similarity: given the embedding of a question, image, or product, they return the stored items whose vectors are closest in meaning.
They use approximate nearest neighbor algorithms, such as HNSW graphs, to search millions or billions of vectors quickly, trading a small amount of accuracy for large speed gains. Vector databases are a core component of retrieval-augmented generation, semantic search, recommendation systems, and image similarity search. Options include dedicated systems and vector extensions to existing databases, such as pgvector for PostgreSQL. Practical considerations include filtering by metadata, keeping embeddings up to date, and managing memory costs.
Related terms
Embedding
A list of numbers (a vector) that represents a word, sentence, image, or other item so that similar items have similar vectors.
NoSQLA broad category of databases that store data in formats other than traditional relational tables, such as documents, key-value pairs, or graphs.
Retrieval-Augmented GenerationA technique that improves AI answers by first retrieving relevant documents and then giving them to a language model as context.
Semantic SearchSearch that finds results based on the meaning of a query rather than exact keyword matches.
Guides that mention Vector Database
Structured vs Unstructured Data: Differences and Examples
What structured, unstructured and semi-structured data are, with real business examples, a side-by-side comparison table and how AI is turning unused documents and recordings into usable data.
Artificial Intelligence
Trending
What Is RAG (Retrieval-Augmented Generation)? How It Works and When to Use It
RAG lets an AI model answer from your own documents instead of guessing. Here is how it works, when to choose it over fine-tuning, and the pitfalls to avoid.