Harness 2023 Predictions: The Year of Automation, Platform Engineering, and More

As the economy tightened during 2022, we've seen the technology industry under increased pressure to maintain growth and innovation. The DevOps landscape is shifting as a result of this urgency, with companies turning toward automation and platform engineering to achieve their goals in an efficient and sustainable way.

Heading into 2023, we will see companies lean heavily into these platforms and advocate more for the developer experience in order to create great products and foster highly collaborative teams. Here are four of my top predictions related to DevOps, GitOps, automation, and more in 2023.

Continued adoption of AI/ML to ease developer burnout

In 2023, we're going to see the adoption of AI and ML increase exponentially. Considering the developer burnout our industry is facing, teams have been hastening their AI and ML adoption - and they're finding that the technology enables faster and more efficient software delivery.

Developer burnout has become one of the most pressing pain points facing the DevOps industry. Developer teams want to continue running at break-neck speed but are often tripped up by the worst parts of the job. They're learning that the main way to keep pace is to automate those parts by leveraging AI and ML. By doing so, they've been able to scale down the burden of tasks such as security scan aggregation and deduplication, deployment verification, reducing the number of tests that are necessary to run on code check-in, and the instantiation and tear down of systems when not in use. Automating these tasks  dramatically reduces toil throughout the entire software delivery lifecycle. In turn, developers get precious time back, which allows them to unleash their creativity and spark new innovation for their organizations.

The rise of platform engineering

We're going to see a wholesale shift from DevOps to platform engineering in 2023. Historically, the terms "DevOps" or "DevSecOps" have tried to include too much in their job function, so we are now seeing developer teams increasingly adopt platforms that streamline tasks such as developing rules and policies,  creating pipelines, and writing code. The industry is going to lean heavily into the adoption of platforms that empower teams to do their best work by not overburdening one group with certain tasks and, instead, operating as a team in a platform. By leveraging a platform engineering approach, developer teams can  work smarter, faster, and provide more business value with stronger outcomes.

Measuring developer effectiveness

Additionally, in 2023, I predict we will see a major shift in how businesses measure the effectiveness of developers' work. I believe that companies will start analyzing developer activities and outputs, similar to how sales teams are evaluated, and that an element of gamification may come into play as well.

With businesses now able to access critical tools that measure employee performance across departments, developer teams will be able to showcase the invaluable work they are doing, and how they are achieving those outcomes. I think this will lead to a positive shift in the way businesses run their tasks and  teams because it will advocate for the most critical facets of the company, like engineering and development.

GitOps as the next generation of infrastructure provisioning

We used to heavily lean on provisioners like SaltStack, Ansible, Chef, and Puppet to ensure all facets of infrastructure were running correctly. GitOps focuses on using git repositories to manage infrastructure and application code deployments and is defined around what is there, and knowing what is ok for  infrastructure and the operations teams. GitOps, as it is now within the Kubernetes world, focuses on the what: figuring out what is running.

But knowing just what isn't enough for audit, compliance, and security in 2023: the "how" is more important. I think there will be an expansion beyond Kubernetes and we will see companies trying to leverage a tool that's defined for the "what" and try to make those systems do the "how" - which they're going to  struggle with.

As a result, companies will need to implement different platforms, execution engines, and pipeline tools to help them figure out the "how." We'll see a significant change in companies focusing on how things are getting into production, not just what's running in production with GitOps.