Azure SRE Agent Architecture and Creation: Practical Benefits for SAP on Azure Customers

AnuradhaKarnam

Microsoft

Feb 26, 2026

Introduction:

SRE Agent is an AI-powered service designed to support site reliability engineering practices through automation and intelligent decision-making. It reduces operational toil, improves uptime, and delivers consistent results by seamlessly integrating with Azure services and external systems to perform operational tasks with limited manual intervention.

Azure SRE Agent reduces operational toil by automating routine and repetitive tasks, allowing teams to concentrate on high impact initiatives. Operational work frequently involves managing diverse Azure resources in combination with on-premises environments, often requiring orchestration across multiple tools. SRE Agent delivers an AI driven platform that unifies these systems and automates operational workflows from start to finish.

How Azure SRE Agent Architecture and Creation Benefit SAP on Azure Customers:

The SRE Agent architecture is particularly well suited for SAP workloads, which are inherently mission critical and span multiple Azure services, including compute, storage, networking, databases, and monitoring. By creating an Azure SRE Agent and associating it with SAP related resource groups, customers gain a unified operational control plane that continuously analyzes telemetry from Azure Monitor, logs, and metrics to identify issues impacting SAP availability, performance, and stability.

Through automated diagnostics, root cause analysis, and guided or approval-based remediation, Azure SRE Agent significantly reduces manual troubleshooting during SAP incidents. In addition, its support for scheduled health checks, configuration validation, and compliance audits aligns closely with SAP best practices and change controlled environments, enabling customers to transition from reactive operations to a proactive, automated, and scalable model that improves uptime and operational confidence at scale.

Centralized Azure Service Management Capabilities:

This diagram illustrates Azure SRE Agent as the centralized automation and intelligence layer that manages Azure resources through Azure CLI and REST APIs, providing a unified control plane for operational tasks across the platform. From this single point, the agent connects to five core service domains: Compute (such as Virtual Machines, App Service, Container Apps, AKS, Functions and more), Storage (including Blob storage, file shares, managed disks, and storage accounts), Networking (covering Vnets, load balancers, application gateways, and network security groups), Databases (Azure SQL, Cosmos DB, PostgreSQL, MySQL, and Redis), and Monitoring & Management (Azure Monitor, Log Analytics, Application Insights, and Azure Resource Manager). Together, the layout shows how Azure SRE Agent enables consistent, automated, and scalable operations across diverse Azure services from a single, AI-driven management layer.

Creating an SRE Agent in the Azure Portal:

Access the Azure portal and complete the following steps to create an SRE Agent.

From Home → Create a resource, search for “sre agent” in the Azure Marketplace. The results clearly highlight Azure SRE Agent (Preview) as an official Microsoft Azure service, confirming that it is provisioned like any other native Azure resource rather than an external tool or addon.

Please select the Azure subscription in which the agent will be deployed and confirm the available Azure SRE Agent (Preview) plan.

In the Basics step, select the subscription and resource group where the Azure SRE Agent will be created. You then provide agent specific details, including the agent's name and the Azure region in which the agent will be deployed and operated. This configuration ensures that the SRE Agent is established as a first-class Azure resource, governed, scoped, and managed using the same Azure constructs as any other native service.

In this step, you define the level of access the agent will have over the Azure resource groups it manages, ensuring alignment with your organization’s security and governance requirements.

Two permission levels are available:

Reader: The agent has read only access to the assigned resource groups. It can observe resource state, analyze telemetry, and generate insights, but any remediation actions require temporary elevation using the user’s permissions after explicit approval.

Privileged: The agent is granted permission to execute approved actions directly on detected resources and resource types within its assigned resource groups. This enables faster, more automated remediation while still operating within Azure RBAC controls and approval of workflows.

This screen confirms that the Azure SRE Agent (Preview) has been successfully deployed in the Azure Portal. The banner “Your deployment is complete” indicates that all required resources were provisioned without errors and that the agent is now active in the selected subscription and resource group.

This screen shows the Azure portal search experience after the Azure SRE Agent has been successfully deployed. By typing “my” in the top search bar, the portal surfaces both services and resources associated with the user’s subscription. Under the Resources section, the newly created Azure SRE Agent instance (for example, mysreapp) appears, confirming that the agent is now registered.

Below screen shows the Azure SRE Agent chat interface for the deployed agent (mysreapp) within the Azure portal. It represents the primary interaction surface where users engage with Azure SRE Agent using natural language to monitor, diagnose, and remediate issues across the Azure resources associated with the agent.

On the left navigation pane, users can manage chat threads, review activities, access the agent builder, monitor health and insights, and configure settings. The main panel displays a new chat thread with a prompt inviting the user to ask a question or execute a command. The quick action buttons (such as App Services, Container Apps and AKS) provide guided entry points to common operational scenarios, helping users get started quickly without needing to remember specific commands.

Once the chat window opens, Azure SAP customers can begin interacting with the Azure SRE Agent using natural language to monitor and manage their SAP landscapes on Azure. To get started, try questions such as:

What can you help me with my SAP systems?

Which SAP subscriptions, resource groups, or SAP related resources are you managing?

What alerts should I configure for my SAP workload (for example, SAP HANA, ASCS, or application servers)?

Show me a comparison of successful requests versus errors for SAP dependent applications across all subscriptions.

If you are troubleshooting a specific SAP issue, you can ask more targeted questions, for example:

Why is my SAP system or SAP HANA database slow?

Why is my SAP application or central services instance not responding?

Can you investigate issues with my SAP workload?

Can you retrieve key metrics (such as CPU, memory, disk I/O, or HANA latency) for my SAP resources?

Conclusion:

Azure SRE Agent empowers SAP customers with a centralized, AI driven operations layer built for managing complex, SAP landscapes on Azure. By integrating natively with Azure and using standard management interfaces, the agent delivers continuous, end-to-end visibility across the compute, storage, networking, database, and monitoring layers that underpin SAP workloads. This unified operational view enables teams to detect and understand issues affecting SAP availability, performance, and stability faster and with greater confidence.

By combining automated diagnostics, intelligent root cause analysis, and guided or approval-based remediation, Azure SRE Agent dramatically reduces manual effort and accelerates incident resolution. Built-in support for proactive health checks, configuration validation, and compliance auditing aligns with SAP best practices and change controlled environments, allowing customers to move beyond reactive firefighting.