2025 DevOps Trends: What Will Change and What Won't

By Derek Ashmore, Asperitas

'Tis the season for issuing predictions about DevOps tool and technology trends that will predominate in the new year.

Unlike your typical prognosticator, however, I don't think that the key to understanding 2025 trends is to focus only on what's new and novel. It's equally important to predict how DevOps won't change over the coming year.

After all, DevOps is a pretty well-established and mature concept at this point. Although there remain opportunities for the typical business to improve upon DevOps practices, it's also the case that many of the DevOps investments that organizations already have in place will continue to deliver value.

To that end, here's a look at eight DevOps trends to watch in the new year. As I explain, some of these trends are new, while others will already be familiar.

New and Growing DevOps Trends to Watch in 2025

Let's start with DevOps trends that remain novel enough to be innovative for most companies. These are examples of DevOps tools or practices that the typical organization doesn't already have in place but that represent opportunities for deriving greater value from DevOps.

1. GitOps

GitOps — which uses Git as a central source of truth for managing infrastructure and application configurations — offers a variety of benefits. It provides centralized visibility into IT resources, the ability to automate provisioning, and simplified automated testing (because teams can run tests against the code stored in Git repositories).

Related:Comprehensive Guide to What a DevOps Engineer Does

GitOps has been around since 2016, but it has only become a prominent DevOps practice recently. Heading into 2025, I see room for GitOps adoption to continue growing as more and more organizations seize upon this valuable DevOps technique.

2. Platform Engineering

Platform engineering, which provides a company's developers with preconfigured tools and services that they can use to build applications, is another DevOps trend that's not entirely new but has plenty of room to expand in the typical organization.

In 2025 and beyond, I foresee platform engineering adoption continuing as more companies look for ways to accelerate application development by giving software engineering teams the car instead of a collection of car parts, so to speak.

3. Infrastructure Static Code Analysis

As development teams move faster, they also need to keep security risks in check. This is where practices like infrastructure static code analysis come into play. Static code analysis provides an automated, scalable way of detecting security risks early in the development process.

Related:The Downside of DevOps: Stress and Burnout of Engineers

4. AIOps

AIOps — by which I mean the IT operations and services necessary to support the design, development, testing, and deployment of AI tools and services — is another DevOps trend that I think will dominate 2025. The reason why is simple in this case: As companies move from talking about AI to implementing it, they'll need systematic operational practices in place to support mature AI products.

Established but Continuing DevOps Trends

Alongside the relatively novel DevOps trends and practices I just described, there is also a set of established DevOps practices that will remain essential to "doing" DevOps successfully in 2025.

5. Policy-Based Governance

Policy-based governance is the practice of defining rules that IT resources must meet using code, then automatically assessing whether the resources align with the rules. Many organizations already have some form of policy-based governance in place, especially if they've been using the cloud for any length of time. As IT environments continue to grow in scale and complexity, policy-based governance will remain critical for meeting security and compliance mandates.

6. SaaS

Software as a service (SaaS) became popular because SaaS deployments reduce the burden placed on DevOps teams to maintain applications and their underlying infrastructure. In this respect, SaaS remains just as valuable today as it was when the SaaS concept became popular two decades ago.

Expect SaaS to remain the preferred model for deploying commercial off-the-shelf software (COTS) in 2025 and beyond, and to see more organizations substituting SaaS for apps that they would otherwise have to deploy and manage themselves.

7. Cloud Migration

In a similar vein, the cloud continues to provide enormous value by allowing DevOps teams to exit the "property management" business — by which I mean the business of having to deploy and maintain physical infrastructure.

We'll continue to see migration of workloads to the cloud in the coming year as more and more organizations work toward an IT strategy in which virtually everything is cloud-based.

8. Serverless

Serverless is an example of a type of cloud service that doubles down on the value of the cloud in general. The reason why is that with serverless, not only do DevOps engineers not have to manage a physical server, but they also don't have to provision or monitor any kind of operating system environment. They simply deploy applications as serverless functions.

While not all apps are good candidates for a serverless approach, expect to see more and more organizations taking advantage of serverless for use cases where it's appropriate in 2025.

Bonus DevOps Trends: DevOps Practices That Will Decline in 2025

In addition to the new and existing DevOps trends that I expect to define 2025, there are certain longstanding DevOps practices that are likely to see a decline in the coming year.

1. Virtual Machine Usage

One is reliance on virtual machines (VMs). As I just mentioned, DevOps teams today don't want to have to maintain any kind of server environment — even a software-defined one — when doing so is not necessary. VMs won't go away, but their usage will decline as teams opt for serverless options wherever possible.

2. Configuration Management Tooling

As use of VMs declines, so too will reliance on configuration management tools designed to automate the provisioning of VMs. Here again, don't expect these tools to go away entirely, but expect them to become less central to DevOps tool sets than they were historically.

Conclusion

In short, as DevOps teams look for even more ways to automate and scale processes, and as they contend with ever-increasing complexity, they'll turn to new tools and practices (like platform engineering and GitOps) in some cases, while also doubling down on existing techniques (like serverless). And they just might shut down some VMs in the process.

About the author:

Derek Ashmore is Application Transformation Principal at Asperitas.