Agentic DevOps: Evolving software development with GitHub Copilot and Microsoft Azure
As someone who’s spent years helping teams navigate the complexities of modern software delivery, I’ve seen firsthand how the developer experience has evolved—and how it’s often weighed down by technical debt, fragmented workflows, and the grind of maintenance. But at Microsoft Build 2025, something truly transformative was unveiled: Agentic DevOps.
At Microsoft Build 2025, we witnessed a pivotal moment in the evolution of software development. GitHub Copilot has introduced a powerful new coding agent—designed to work alongside developers from ideation to deployment. This marks a transformative shift in how we approach the entire software development lifecycle, with intelligent agents now playing a central role—and developers firmly in control.
Just weeks ago, as Microsoft celebrated its 50th anniversary, we revisited the original vision of a “software factory.” It’s a concept that has always resonated with me: a place where innovation is built by developers, for developers. Now, as we step into Microsoft’s fifty-first year, that vision is being redefined. The software factory of the future is intelligent, collaborative, and developer-centric—and GitHub Copilot is leading the way.
Being a developer often feels like being a modern-day wizard—turning ideas into reality with nothing but code. But as systems scale and complexity grows, the joy of building can get buried under mounting technical debt, endless maintenance, and constant context switching. The magic fades, and momentum stalls.
GitHub Copilot has already revolutionized how we write code, helping over 15 million developers build faster and smarter. But coding is just one part of the journey. Our days are filled with designing architectures, debugging, refactoring, navigating documentation, and wrangling legacy systems.
Now, with agentic workflows powered by GitHub Copilot, we’re entering a new phase—one where intelligent agents assist across the entire development lifecycle. From accelerating design and boosting code quality to enhancing collaboration, improving security, and even managing technical debt in production, these workflows are helping us move from idea to impact faster than ever.
This isn’t just about productivity—it’s about bringing back the flow, the creativity, and the joy of building great software.
Introducing agentic DevOps
We’re entering a new chapter in software development—one where intelligent agents don’t just assist, they collaborate. This is the foundation of a new approach we’re calling Agentic DevOps: the next evolution of DevOps, reimagined for a world where AI-powered agents work alongside developers and with each other to automate and optimize every stage of the software lifecycle.
Think of it as adding a crew of tireless teammates to your development squad. These agents handle bug fixes, implement small features, maintain documentation, and more—freeing you up to focus on the work that truly moves the needle.
In the traditional model, software development often felt like a slow relay—weeks to plan, months to build, quarters to launch. But today, ideas can become prototypes in hours and reach production in days. Agentic DevOps is designed to help you build faster, crush your backlog, eliminate technical debt, secure your applications, and keep everything running smoothly in production.
And the best part? You remain at the center of it all—conducting this intelligent orchestra, approving recommendations, and driving innovation. It’s about putting developers back in control, so we can get back to building epic things.
Let’s face it—development can sometimes feel like a grind. The creative spark that brought many of us into this industry can get buried under repetitive tasks, technical debt, and endless debugging. GitHub Copilot is changing that. It’s not just about code completion anymore—it’s about reclaiming the joy of building.
Copilot is already helping developers write better code, surface vulnerabilities, and resolve issues before they hit production. And now, it’s evolving to power a new paradigm: Agentic DevOps. Here’s what’s new and how it all fits together:
🚀 Agent Mode
Agent Mode takes everything we love about GitHub Copilot and supercharges it. It now supports complex, multistep coding tasks in a seamless, collaborative experience. From analyzing entire codebases and editing across files to generating tests, fixing bugs, and suggesting terminal commands—all from a single prompt. And it’s available in all major editors: Visual Studio Code, Visual Studio, JetBrains, Eclipse, and Xcode.
🤖 New Coding Agent
The latest coding agent transforms Copilot from a helpful assistant into a true peer. It can take on tasks like code reviews, writing tests, fixing bugs, and even implementing full specs. It collaborates with other agents to tackle complex challenges across development and production. With built-in audit logs and branch protections, every change is reviewed before it ships—making Copilot a trusted teammate.
Recommended by LinkedIn
🧱 App Modernization Capabilities
Shipping to production isn’t the end—it’s just the beginning. Copilot now helps modernize legacy stacks, handling code assessments, dependency updates, and remediation for Java and .NET apps. Mainframe support is on the horizon. It generates and executes update plans automatically, giving you full control and visibility. The result? More secure, maintainable, and cost-efficient apps—and a clear path to innovation.
🌐 Azure SRE Agent
Introducing the new Site Reliability Engineering Agent for Azure. It monitors production systems 24/7, responds to incidents in real time, and troubleshoots autonomously. Whether you're running AKS, App Service, serverless, or databases, this agent keeps your systems healthy and resilient. It logs remediation actions as GitHub issues, ensuring your team can follow through and close the loop. Fewer wake-up calls, faster recovery, and more time to build what’s next.
Beyond the exciting agentic capabilities we’re rolling out, there’s another major leap forward: GitHub Models. If GitHub Copilot changed how we write code, Azure AI Foundry is changing what we can build—and now, that power is seamlessly integrated into your GitHub workflow.
With this new native integration, developers can experiment with cutting-edge models from OpenAI, Meta, Microsoft, Mistral, Cohere, and more—right from GitHub. You can:
And with built-in enterprise-grade guardrails, your model choices remain secure, responsible, and aligned with your organization’s policies.
This is a game-changer for AI app development—bringing experimentation, deployment, and governance into one streamlined experience. Whether you're building intelligent agents or modernizing legacy systems, GitHub Models makes it easier to innovate with confidence.
If you would like to know more about these SRE agents read my past article on Azure SRE Agents.
One More Thing That Changes the Game: GitHub Copilot Goes Open Source in VS Code
As GitHub Copilot continues to evolve within Visual Studio Code, AI has become a core part of the developer experience—enhancing everything from testing to security. Now, we’re taking a bold step forward in our open-source journey: Microsoft is open-sourcing GitHub Copilot in VS Code.
This means the AI-powered capabilities from the GitHub Copilot extensions will now be part of the same open-source repository that powers the world’s most popular development environment. It’s a move that reflects our deep commitment to:
By opening up Copilot’s integration in VS Code, we’re giving developers greater visibility, more control, and a stronger voice in how AI evolves within their workflows. It’s not just about building smarter tools—it’s about building them together.
This isn’t just another buzzword. It’s a reimagining of the entire software development lifecycle—powered by intelligent agents that work with you, not just for you. Microsoft and GitHub are leading this shift, and I believe it’s going to redefine how we build, collaborate, and innovate.
This is the future of DevOps—agentic, intelligent, and developer-first. And you’re still at the center, orchestrating it all.
Cheers! 😎
