EDB Postgres AI

EDB Postgres AI (EDB PG AI) is a sovereign data-and-AI platform built on open-source PostgreSQL that unifies transactional, analytical, vector, and agentic workloads through a single SQL interface. It is aimed at CIOs, CTOs, and data leaders who need AI-ready infrastructure without moving data into a separate cloud or vector store.

Announced on June 23, 2026, EDB Postgres AI is EnterpriseDB's platform for bringing AI to where enterprise data already lives, built on open-source PostgreSQL to preserve data sovereignty. Its new agentic-database capability turns the database into a self-optimizing system that monitors 200+ operational and performance metrics and automates tuning (EDB cites up to 10x faster tuning and up to 8x faster application performance), while unifying relational, JSON, time-series, geospatial, and vector data through a single SQL interface with row-level and role-based access controls. Its converged-analytics capability uses a zero-ETL architecture that publishes operational data to Apache Iceberg and adds EDB PG AI for ClickHouse (real-time) and WarehousePG (historical) to span real-time to petabyte-scale analytics without separate extract-load pipelines. Native vector search lets AI agents retrieve accurate context from data they are already authorized to access, eliminating a separate vector store to secure and synchronize. A governance-at-the-data-layer capability, which enforces rules row-by-row and gives every agent a declared owner and boundary, is in preview with GA expected in the second half of 2026. Partners include Dell, IBM, NVIDIA, Red Hat, and Supermicro.

Category

Data & Analytics

Pricing

Subscription, Contact for pricing

Target Market

CIOs, CTOs, Data Engineers, Database Administrators, Enterprise Architects

Founded

2004

Headquarters

Wilmington, Delaware, USA

• ✓
  Agentic database

  A self-optimizing system that monitors 200+ metrics and automates tuning, with EDB citing up to 10x faster tuning and up to 8x faster application performance.

• ✓
  Converged analytics

  A zero-ETL architecture that publishes operational data to Apache Iceberg and adds ClickHouse and WarehousePG for real-time to petabyte-scale analytics.

• ✓
  Native vector search

  A unified query layer over vector, structured, and unstructured data so agents retrieve context they are authorized to access without a separate vector store.

• ✓
  Governance at the data layer

  Row-level and role-based access enforced in the moment of action, giving every agent a declared owner and boundary (in preview, GA expected H2 2026).

• ✓
  Sovereign, open-source Postgres foundation

  Built on open-source PostgreSQL and deployable on-premises, hybrid, or across multiple clouds to preserve data sovereignty.

• •
  Unify AI and operational data

  Serve relational, JSON, time-series, geospatial, and vector data through a single SQL interface for AI applications.

• •
  Reduce database sprawl

  Consolidate transactional, analytical, and vector workloads onto one governed Postgres platform instead of many specialized stores.

• •
  Ground enterprise AI agents

  Give agents accurate, authorized retrieval on live data with governance enforced at the row level.

EDB references predictable per-core pricing and offers bring-your-own-cloud (BYOC) and an EDB Developer Cloud; detailed list pricing is not published in the announcement.

This page was written from 2 sources, 2 on domains other than enterprisedb.com.

1. 1.prnewswire.com — edb launches agentic database converged analytics and govern
2. 2.techtarget.com — Latest EnterpriseDB features unify data for AI development