Artificial intelligence is transforming the way software is built, deployed, and operated. AI-assisted tools are now a familiar part of the development lifecycle, with teams racing to use them to reap benefits such as accelerated code generation, improved test coverage, and better observability and incident response.
But increasing velocity introduces a new challenge: reliability. Modern software delivery depends on distributed systems, streaming pipelines, and large volumes of operational data. If that data is inconsistent or poorly governed, the impact can spread across entire development and delivery pipelines.
These challenges were explored in the recent DevOps in the AI Era: Fueling Innovation at Scale roundtable hosted by Database Trends and Applications (DBTA). A key focus of the discussion was how AI for DevOps is transforming expectations, making data reliability vital to modern software delivery.
The panel featured our CEO Peter Caron, alongside Jatinder Luthra, Sales Engineering Advisor at Perforce, Kellyn Gorman, Multiplatform Database and AI Engineer and Advocate at Redgate and was moderated by Stephen Faig, Research Director at Unisphere Research and Database Trends and Applications (DBTA).
Here are some of the key takeaways:
AI adoption requires more than task automation
During his presentation, our CEO Peter Caron challenged a common assumption about AI adoption. Many organizations view AI primarily as a way to automate tasks or reduce engineering costs. But Peter pointed out that it misses the broader structural impact AI will have on organizations.
“Adopting AI in an organization is not a question of simply automating tasks or using smarter tools. Those alone are insufficient and cannot be the end objective,” Peter said.
Instead, he believes companies must rethink how work is organized.
“Companies will need to redesign existing jobs and change roles to be broader and more flexible if they’re going to succeed in this next technological disruption.”
In other words, AI is not just improving individual tasks, it’s reshaping the systems the tasks exist in.
AI won’t replace developers, but it will reshape their roles
Peter also commented on the idea that generative AI tools will replace software engineers entirely. While AI coding tools are becoming increasingly sophisticated, he warned this oversimplifies how complex engineering systems actually function.
“For every complex problem, there is an answer that is clear, simple, and wrong,” he said.
Peter explained that replacing developers with AI agents may seem like an obvious efficiency gain, but it misunderstands how workflows operate across teams, tools, and systems. Even when AI can automate part of a workflow, the broader system still requires coordination, judgment, and oversight.
“Using AI to build tools in isolation will not always result in an improved system,” he added.
Instead of eliminating engineers, AI is far more likely to change how engineering roles are structured.
The biggest impact of AI will happen at the system level
One of the key ideas in Peter’s presentation was the distinction between optimizing tasks and redesigning systems. He brought this to life with a historical example. When word processors first appeared in the late 1970s, the role of typists largely disappeared. But typing itself did not disappear. The task simply moved to other roles within organizations, and the system around the task changed. The same principle applies to AI-driven development today.
Peter said: “The real impact of AI comes not from how it performs a task, but from how it restructures the entire system around that task.”
As AI becomes integrated into development workflows, engineers, DevOps professionals, and platform teams will need to adapt to new responsibilities.
DevOps roles will evolve as AI changes how work is organized
Peter believes the most successful organizations will redesign workflows around AI capabilities, rather than just inserting AI into existing ones. Engineers and DevOps professionals will increasingly focus on orchestrating systems, validating outputs, managing constraints, and ensuring the reliability of the broader environment. Over time, the tasks that make up roles such as software engineer, DevOps engineer, and database engineer will evolve significantly.
“The great AI swap is not coming for everyone’s job. But the way we use AI will certainly change what the definition of a job is,” Peter said.
“We will unbundle tasks and rearrange them into new roles so that the job titles remain, but the tasks inside those roles will be very different in the next few years.”
Data visibility and governance are essential for AI-driven development
As AI becomes more deeply embedded in development workflows, organizations must also ensure the data underlying those systems remains reliable and understandable.
Without visibility into operational data, schema changes, or access controls, teams can quickly lose confidence in the systems they depend on. This is where stronger data tooling and governance become essential.
Solutions like Studio 3T help organizations working with MongoDB improve visibility into their data environments, control access across teams, and safely manage schema changes and operational data. In AI-driven development environments, this visibility is essential for maintaining both velocity and trust.
Watch the webinar on demand
The full conversation offers valuable insights for engineering leaders, DevOps practitioners, and data teams navigating the intersection of AI, automation, and modern software delivery.
Watch the full discussion here:
https://www.dbta.com/Webinars/2348-DevOps-in-the-AI-Era-Fueling-Innovation-at-Scale.htm?src=3ts
One message from the panel was clear: the organizations that succeed in the AI era will not simply automate tasks. They will rethink the systems, roles, and data foundations that power modern software development.
