⚡
  • HOME
  • NEWS
  • SERVICES
  • ARCHITECTURE
  • TECH STACK
  • PORTFOLIO
  • ABOUT
  • CONTACT
HOMENEWSSERVICESARCHITECTURETECH STACKPORTFOLIOABOUTCONTACT
© 2026 Miodrag Gromilić. All rights reserved.
HOMENEWSSERVICESTECH STACKPORTFOLIOCONTACTABOUTFAQs
Back to Skills

pgvector

Since 2021Vector search as a PostgreSQL extension

Vector search as a PostgreSQL extension — simplest path to AI search without another database.

Overview

Game-changer — vector similarity directly in PostgreSQL. Embeddings alongside relational data, one DB, one backup. HNSW indexes for fast ANN search, combined with full-text search and JSONB in single queries.

Use Cases

  • AI search in PostgreSQL
  • RAG without extra infra
  • hybrid relational + vector queries
  • embedding storage

Advantages

  • No extra DB to manage
  • full SQL alongside vectors
  • HNSW indexing
  • transactional consistency
  • simple install

Considerations

  • May not match dedicated vector DBs at extreme scale
  • PostgreSQL only
  • evolving features

Works Great With

PostgreSQLPythonLangChainLaravelNode.jsOpenAIDocker

Related Technologies

PostgreSQL
Since 1996
MySQL
Since 1995
Redis
Since 2009
SQLite
Since 2000
MongoDB
Since 2009
Elasticsearch
Since 2010
Back to Skills CONTACT