Azure/AI-Landing-Zones

What it solves

This project provides a production-ready reference architecture for deploying secure and resilient AI applications and agents on Azure. It eliminates the need to build the foundational infrastructure from scratch by providing a standardized, enterprise-scale foundation for both generative and non-generative AI workloads.

How it works

The project provides Infrastructure-as-Code (IaC) implementations using Bicep and Terraform, based on Azure Verified Modules. It is divided into two primary components:

• AI Foundry Landing Zone: Leverages Azure AI Foundry for AI apps and agents workloads.
• AI Gateway Landing Zone: Uses Azure API Management (APIM) as an AI Gateway to centrally manage and serve Foundry models.

These components can be deployed together or independently depending on the organization's needs.

Who it’s for

Enterprise architects and DevOps engineers who need to deploy AI workloads on Azure while adhering to the Cloud Adoption Framework (CAF) and Well-Architected Framework (WAF).

Highlights

• Enterprise-Scale: Designed for production-ready, secure, and resilient AI environments.
• Multi-Implementation: Available via Terraform, Bicep, and the Azure Portal.
• Flexible Architecture: Supports both generative and non-generative AI scenarios.
• Comprehensive Guidance: Includes a design checklist covering CAF and WAF design areas to help assess greenfield and brownfield environments.