Agentic DevOps
Agentic DevOps: Using Microsoft Azure and GitHub Copilot to advance software development
Microsoft Reveals the Next Development in Software Development with Agentic DevOps At Build 2025
Microsoft today announced the launch of Agentic DevOps, a novel strategy that uses intelligent agents to rethink the creation and management of applications, marking a critical milestone in the evolution of the software development lifecycle. Known as the “next evolution of DevOps,” Agentic DevOps imagines a future in which AI agents work together with developers and with one another to automate and optimize the whole software lifecycle.
As it commemorates its 51st anniversary and considers how it began as a “software factory” created by developers for developers, the company is rethinking that factory with the developer at its core. The goal is to assist developers in “breaking free from the grind,” regaining the “joy, your flow, and the magic of building” that can occasionally be lost in the face of mounting demands, complexity, and technological debt. Although a developer’s work includes coding, they spend a large amount of their day developing systems, sifting through documentation, troubleshooting, refactoring, and battling legacy code.
GitHub Copilot, which is positioned as essential to this change, introduces agentic workflows. Already used by 15 million developers to speed up development, GitHub Copilot is developing beyond code completion to facilitate these agentic workflows, which help teams move from idea to production more quickly, improve code quality, collaborate more quickly, fortify security, pay off technical debt, and maintain the smooth operation of apps in production.
In an effort to promote openness and community-driven innovation, Microsoft is incorporating its AI-powered features into the Visual Studio Code open-source repository by making GitHub Copilot available to all users.
New specialized agents made to work with developers are a major component of the Agentic DevOps announcement. GitHub Copilot gets promoted from pair programming to “peer” status in the development team with the introduction of a new coding agent. This agent can work with other agents and manage intricate, multi-step coding jobs. Code reviews, test authoring, issue fixes, and full specification implementation are just a few of the duties developers might give it.
It can even recommend terminal commands from a single prompt and is made to function across files. In well-known code editors like Microsoft Visual Studio Code, Visual Studio, JetBrains, Eclipse, and Xcode, this “Agent mode” is accessible. Before being deployed, each suggested modification from this agent must be reviewed because it has integrated audit logs and branch protections.
The introduction of a new Azure Site Reliability Engineering (SRE) Agent goes beyond the development stage. By keeping an eye on production systems 24/7, this agent relieves developers of the burden of late-night warnings. It automatically diagnoses and responds to issues. Using Microsoft’s global experience running Azure-based services, the SRE Agent checks application health and performance in Azure Kubernetes Service (AKS), Azure App Service, serverless, and databases.
When problems arise, it may either solve them on its own or interactively lead developers through the root cause analysis. Importantly, any repair items or remediation actions are recorded as GitHub issues so that the team may follow up on them. The objective is more resilient systems that can self-heal, allowing teams to remain relaxed and focused, quicker recovery, and fewer wake-up calls.
Technical debt is one of the enduring issues that Agentic DevOps is intended to address. With mainframe modernization scheduled for the near future, GitHub Copilot is launching new app modernization features to assist developers with updating stacks, with a focus on older Java and.NET apps. Code assessments, dependency updates, and remediation are aided by these features, which also give visibility and control over the changes while automatically creating and carrying out update plans. The goal is to create applications that are more cost-effective, safe, and stable so that developers may concentrate on innovation rather than being constrained by the past.
Microsoft Azure services, such as those for AI + machine learning (like Azure AI Foundry and Azure OpenAI in Foundry Models), containers (Azure Kubernetes Service, Azure App Service), databases (Azure Cosmos DB, Azure DB for PostgreSQL), and monitoring (Azure Monitor), comprise the underlying platform that supports these capabilities. Azure AI Foundry is marketed as an “AI App and agent factory” that makes it possible to test out state-of-the-art models from several suppliers, all of which are easily available within the GitHub workflow through a new native integration. This integration makes it possible to invoke models or agents from GitHub Actions, compare models, and exchange them using a single API.
According to Microsoft, Agentic DevOps is “as transformative as the shift to the cloud” and a “seismic shift.” It promises to rewrite cost structures that have limited teams for decades, eliminate friction, and simplify processes. Agentic DevOps seeks to not only boost productivity but also unleash the potential for creating the future by automating repetitive chores and allowing developers to concentrate on innovative, high-impact work.
Through presentations devoted to reinventing software development with agentic AI, collaborative development with GitHub Copilot, accelerating Azure development, app modernization with AI, and agent mode in action, attendees of Microsoft Build 2025 are encouraged to delve further into this future.
