Guide & Getting Started
Begin your journey with EDB Postgres® AI by exploring how to deploy, manage, and build intelligent systems across hybrid environments using Postgres, GenAI workloads, and analytics pipelines.
What do you want to do?
Do you want to run analytical queries on your Postgres data? Check out the EDB Postgres® AI Analytics documentation, which covers the Lakehouse engine, Delta Lake, Iceberg, and tiered storage.
Do you want to run GenAI or machine learning models on your data? Explore the EDB Postgres® AI Factory, including tools for vector retrieval, agent workflows, and model orchestration.
Do you want to manage hybrid workloads, self-managed databases, or model endpoints? Visit the EDB Postgres® AI Hybrid Manager, which provides a Kubernetes-native control plane.
Hybrid Manager
Hybrid Manager is the control plane that runs EDB Postgres AI. It provides a unified interface for managing database clusters, AI workloads, and analytics engines in any Kubernetes environment.
Start here
- Overview
- Install on Kubernetes
- Manage Postgres clusters
- Observability and diagnostics
- Sovereign AI and Data Factory System
AI Factory
AI Factory is a low-code environment for building GenAI systems, vector-enabled retrieval, and data-connected agents. It runs either standalone or orchestrated through Hybrid Manager.
Learning paths
- 101 - Core concepts and first assistant
- 201 - Integrating knowledge and external data
- 301 - Scaling, tools, and vector retrieval
Analytics & Pipelines
The AI Accelerator module provides Postgres-native analytics using tiered storage, Delta Lake/Iceberg, and vector pipelines.
Start here
Use cases
Recommended next steps
| Area | Start here |
|---|---|
| Hybrid Manager | Hybrid Manager overview |
| AI Factory | 101 Learning Path |
| Analytics | Lakehouse overview |
Once you're set up, explore advanced capabilities:
- Deploy sovereign AI solutions using Hybrid Manager
- Serve and monitor GenAI workloads
- Build tiered data lakes with Delta Lake and Iceberg
- Integrate vector pipelines directly into your database
Could this page be better? Report a problem or suggest an addition!