AI Factory Explained v1.3

Welcome to AI Factory Explained — your deep dive into the architecture, concepts, and components of EDB Postgres® AI (EDB PG AI).

EDB PG AI enables Sovereign AI — building intelligent applications and data products on top of your data, in your databases and data centers, with Postgres as the foundation. The platform integrates vector search, Pipelines, Gen AI Builder, Knowledge Bases, and Model Serving to give you complete control over how AI is embedded in your systems.

This section explains how it all works:

• What key concepts underlie EDB PG AI
• How the architecture fits together
• How individual capabilities like Pipelines, Gen AI Builder, and Model Serving work
• Where vector search and AI models fit within your Postgres estate

Core Concepts

Start here to get oriented to the platform:

• AI Factory concepts overview
• Terminology guide
• Generic concepts

Gen AI and Knowledge Bases

Understand how intelligent knowledge is built on your data:

• Knowledge Bases
• Retrievers
• Rulesets
• Structures
• Threads
• Tools

Data Pipelines and AI Accelerator

Understand how Pipelines transform your AI data workflows:

• Pipelines overview
• Data Lake

Vector Search and PGvector

How vector search is embedded directly in Postgres:

• PGvector and Vector Engine concepts

Model Serving

How AI models are deployed, governed, and used:

• Model Serving concepts
• Model Library (Asset Library) explained

Where to go next:

If you're ready to build:

• Start with a Learning Path
• Follow a How-to Guide

Or explore more advanced solution patterns in Solutions and real-world Use Cases.

This section grows as EDB PG AI evolves — check back for the latest explanations as new capabilities are added.