Azure/AI-Landing-Zones

Azure AI‑Landing‑Zones – Enterprise‑grade reference architecture for AI workloads on Azure

What it is

• A Microsoft‑maintained, production‑ready application landing zone that gives you a ready‑made Azure architecture for deploying AI applications and agents at scale.
• Delivered as Infrastructure‑as‑Code (IaC) in Bicep, Terraform, and a Azure Portal deployment template, all built on the Azure Verified Modules (AVM) library.

Why it matters

• Provides a vetted, secure, and resilient foundation that aligns with the Azure Cloud Adoption Framework (CAF) and the Well‑Architected Framework for AI workloads.
• Handles the full stack of services needed for modern AI (Azure AI Foundry, Azure AI Search, Container Apps, Cosmos DB, Container Registry, App Configuration, Key Vault, zonal compute, etc.) and works in most Azure public regions.

Core components

Component	Purpose	Implementation options
Agent Landing Zone	Sets up the “Foundry” environment where AI apps and autonomous agents run.	Bicep (aka.ms/ailz/bicep), Terraform (aka.ms/ailz/terraform), Portal button
AI Gateway Landing Zone	Provides a centralized API Management (APIM) gateway for exposing and governing AI models.	Bicep (aka.ms/ailz/apim), Terraform (aka.ms/ailz/aigateway/terraform)
Foundry Modules	Reusable building blocks (e.g., AI services, storage, networking) that the two zones consume.	Bicep registry module, Terraform registry module

How you use it

1. Pick a zone – Deploy the Agent zone, the Gateway zone, or both, depending on whether you need just the AI runtime or also a managed API front‑end.
2. Choose IaC flavour – Run the Bicep files with az deployment sub create …, apply the Terraform module, or click the “Deploy to Azure” button in the repository’s portal folder.
3. Customize – Extend the base templates with additional Azure services (e.g., Cognitive Search, custom VMs) to match your specific scenario.
4. Validate – Follow the built‑in Design Checklist (CAF + Well‑Architected) to ensure security, governance, and resiliency.

Typical scenarios covered

• Chat interfaces powered by Azure AI Foundry
• Autonomous agents that call multiple AI services
• Document generation and content processing pipelines
• Knowledge‑mining conversational agents
• Custom “copilot” style assistants
• Modernizing legacy codebases by wrapping them in AI‑enabled APIs

Supported regions (full deployment): australiaeast, eastus, francecentral, germanywestcentral, japaneast, koreacentral, norwayeast, polandcentral, southafricanorth, southeastasia, spaincentral, swedencentral, switzerlandnorth, uaenorth, uksouth.

Status & roadmap

• Currently in preview; roadmap and upcoming features are published at https://aka.ms/ailz/roadmap.
• “What’s New” and FAQ pages keep users informed about preview services and design guidance.

Getting started

# Example: Deploy Agent Landing Zone with Terraform (quick start)
terraform init
terraform apply -var "location=eastus"

Or use the portal button in the repo to launch a guided deployment UI.

Contributing

• Contributions are welcome under Microsoft’s CLA and Open Source Code of Conduct. See the repository’s CONTRIBUTING.md for details.

---

All information above is taken directly from the project’s README; no additional features are inferred.