'Writing new code is just one small part of the job. The rest of a developer's day is packed with designing systems, digging through docs, debugging, refactoring, and wrestling with legacy code.'
At Build 2025, Microsoft unveiled numerous agent-based products and features. However, one tool that may impact software developers in unprecedented ways is Agentic DevOps in GitHub Copilot.
DevOps is a set of practices that unites software development and information technology operations to improve collaboration, automate processes, and deliver software faster and more reliably. Microsoft wants to add an 'agentic' layer, which automates and optimises each stage of the software lifecycle.
"It's like adding a crew of tireless teammates to your developer squad -- handling bug fixes, small features, documentation, and more -- so you can stay focused on the work that matters most," said the company in the blog post.
This is made possible by integrating various agents within GitHub Copilot, which handle coding, app modernisation, site reliability engineering, and other tasks. These agents help developers build faster, reduce backlogs and tech debt, secure their apps, and keep everything running in production.
Microsoft also said, "Writing new code is just one small part of the job. The rest of a developer's day is packed with designing systems, digging through docs, debugging, refactoring, and wrestling with legacy code."
"We want to help you break free from the grind," added Microsoft.
Besides Microsoft, several leaders in the industry echoed the sentiment. "I don't think the next big leap in software productivity comes from 'vibe coding'. It comes from removing the grind in real code bases," said Eric Schmidt, former CEO of Google, about a similar platform on X.
In a conversation with AIM at Build 2025 last week in Seattle, Amanda Silver, corporate VP and head of product at Microsoft's Developer Division, offered further insights. "This [Agentic DevOps] has huge implications for the future of software development in a few different dimensions," said Silver.
"One of the most hated dimensions of modern development is being on-call and getting woken up at 2 am for a live-site incident. Our agentic workflows can handle the first-line of response, help isolate the root cause, and in some cases employ mitigations so you can sleep through the night," she added, indicating the capabilities of the Azure SRE agent, which will then file that as a GitHub issue that will be taken care of by the coding agent.
Silver also said that Agentic DevOps helps developers "tear through crushing technical debt" by automatically submitting fixes for security vulnerabilities it finds and helping modernise codebases, which she claims can save 70% of the manual time.
A recent survey from Harness, a platform that offers agentic capabilities in the software development life cycle, suggests that developers are indeed working on repetitive tasks. An agentic layer can also benefit both the company and its developers.
The study surveyed 500 engineering leaders and practitioners and revealed that 78% of developers spend at least 30% of their time on manual, repetitive tasks. These tasks include writing compliance policies, QA testing, and error remediation.
When developers spend excessive time on the above tasks, which may also lead to overtime, they revealed that it 'creates an unhealthy work/life balance' and increases burnout, stress, and anxiety.
"About 98% of developers believe AI tools are a great way to reduce burnout," the report added. Moreover, the developers also reinforced the dangers and pitfalls in using AI for writing code, and 96% of developers and 95% of engineering leaders observed that the full benefits of AI-assisted software development will 'never be realised' until their use cases extend to the entire software delivery life cycle.
While developers and engineering leaders reinforce the need for a tool like Agentic DevOps, the study also added that companies may see huge cost savings.
"For example, the average salary of the developers we surveyed is $107,599. So if 30% of their job is made up of redundant tasks, that equates to $32,280 of wasted investment in each developer," read the report. It also estimates that organisations with around 250 developers lose at least $8 million in productivity annually.
Silver went back in time and referred to the 'old waterfall model' for software development, in which engineering was ultimately handed off to a validation engineer or a QA engineer to validate that the product was ready to ship.
But when DevOps was introduced, those roles were kind of combined. In the case of Microsoft, the company merged quality assurance and site reliability engineers into a single profession.
Despite this, Silver said that the current model still involves a process where product managers hand off to designers who hand off to engineers, and so on.
"I think what we're going to see [with advancements in AI] is that each of those roles becomes more ambidextrous, and as a result, there's tighter collaboration across all of the teams," said Silver, indicating that the idea of 'waterfall' is going to dissolve.
Silver also suggests a dual approach to AI integration. While it's crucial to integrate AI into existing workflows (meeting developers where they are), she also foresees a future where agents become omnipresent and somewhat detached from specific tools. The intelligence will reside in the knowledge base and context, accessible across various platforms like Visual Studio Code, GitHub, and even communication clients like Microsoft Teams.
This means AI won't just be a feature within an IDE; it will be an intelligent layer that assists developers throughout the entire software development lifecycle, from ideation to operations.
