Modernization remains a priority for CIOs, with IT executives saying that legacy systems and outdated infrastructure .
But the priority today is not just modernization, but modernization at top speed.
“Technology is moving fast, but society and business conditions are moving faster, so now you have to modernize just to keep up,” says veteran IT leader , an IT services and consulting firm.
Researchers, however, aren’t as optimistic, noting that organizations face a host of challenges — , says due to slow decisions and actions. West Monroe calls it the “slowness tax.”
Furthermore, Brown believes AI will play an increasingly more significant role in IT modernization in the years to come. He predicts that AI agents will someday be capable of constantly reviewing the IT environment, including resources within applications, for pieces that are out of date; once agents identify the outdated pieces, they will determine how best to update them and then execute the updates — with limited to no human intervention, only human oversight.
“AI gives us an amazing mechanism to handle continuous modernization,” Brown says, acknowledging agentic AI is not yet capable of doing that yet but is “confident it will be.”
In today’s incarnation, while agentic AI shows promise, , vice president of cloud governance at Capital One, says the company views “modernization as the continuous refinement of our operational DNA, enabling a permanent state of evolution.”
The goal, he explains, “is a frictionless environment where we can rapidly deliver new capabilities, including AI-driven ones, without being slowed down by the weight of how we used to build software and deliver value.”
Case in point is its recent enterprise-wide shift to a serverless approach for application development, which made managed, cloud-native compute as the default for new applications. Sullivan says the change “allowed us to treat infrastructure as a set of services rather than something developers directly operate. We moved away from manual management and toward a model where the platform handles the undifferentiated heavy lifting.”
The shift enabled faster innovation, a stronger security posture, and better cost efficiency, he says.
Given such benefits, Sullivan lists the adoption of managed services and serverless architectures as a key strategy for accelerating modernization.
“CIOs should stop viewing infrastructure as a set of assets to manage and start viewing it as a product delivered through services,” he says. “When you pair this with automated governance and observability, you create a safe space for innovation. It allows teams to move fast without the risk of compromising security or resilience.”
“Teams spend less time on maintenance and more time on the creative part of engineering,” Sullivan adds. “We want them to be able to focus on writing exceptional code and building innovative applications. This reduction in toil translates directly into faster delivery. It creates a culture where experimentation is cheap and quick. When the platform handles the plumbing, the engineers can focus on the product.”
3. Establish a culture of modernization
, a partner with PwC Technology Strategy and leader of the firms’ CIO Program.
To do that, CIOs need to go with modular IT architecture so “everything can be changed out,” she says.
Modular IT architecture features independent, interoperable components that can be developed, deployed, and scaled separately. Because each module has clear interfaces and responsibilities, IT teams can update or replace capabilities without disrupting the entire stack — just as pit crews do when they replace parts but not the racecar itself.
“You want modularity. You don’t want that monolithic model to creep back in. But you also don’t want modularity from the standpoint of fragmentation. The whole key is to create the fewest dependencies, so you can literally unplug something when you want to plug the new thing in,” Phaneuf explains.
This approach is particularly useful when it comes to AI models, Phaneuf says. Ongoing innovations are happening so fast that today’s leading AI model could become tomorrow’s laggard, so organizations need the ability to quickly switch out one model for another when needed.
SOCIAL SHARE CARD GENERATOR