Datacom: Agentic AI overtakes cloud migration as the new modernisation engine
Agentic AI shows promise as a way to tackle lengthy, complex modernisation and create new opportunities for partners.
The market moved from generative AI to agentic AI faster than expected, catching even seasoned technologists by surprise at how quickly maturity arrived, according to Mike Walls, director, cloud A/NZ, Datacom.
Agentic AI is now expected to drive a new wave of tech modernisation thanks to the speed and efficiency of these systems.
It has quickly replaced cloud migration as the primary driver of these initiatives and is opening up long-delayed modernisation projects.
“The problem with a lot of legacy systems is they're very complex, not well understood or well documented, but agentic AI is able to interrogate these systems, understand the code, the architecture and then document the system,” Walls said, speaking at last week’s AWS Re:Invent conference in Las Vegas.
Datacom, a long-term AWS partner, recently signed a five-year strategic collaboration agreement to enhance digital modernisation. It’s seeing an uptick in mainframe modernisation work as organisations consider projects that until now have been deemed too complex and costly.
AI agents can suggest optimal migration paths, after which Datacom can refactor legacy systems using a modern architecture — rewriting the code or re-platforming onto a new database or platform-as-a-service environment.
In some cases, mainframe and other legacy modernisation programs that would usually span a year or more of business analysis, requirements gathering and development, can be compressed them weeks using agentic AI.
“Whereas before business analysts needed to understand and document the legacy system, which was hugely expensive. Then you've got to provide your user requirements, maybe six months to get that all done. Agentic is fast tracking that entire process,” Walls said.
Datacom serves diverse sector, including public, financial and healthcare in Australia and New Zealand, and Walls can point to numerous examples of the transformative impact of AI.
“With payroll, for example, agentic AI can allow people to interact using natural language to understand policy and compliance,” he said.
The agentic AI can access all the relevant systems, data and policies, analyse them and return answers or actions based on that information. But the agentic system has first been put through compliance certification.
“It passed it with flying colours first time,” he noted.
In 2026, Walls expects more AI advancements and the increasing convergence of IT, OT and physical security as OT and security systems are digitised and agentic AI is applied across these environments.
“With agentic AI now being laid into physical and operational environments, it starts to become much more proactive security, rather than just reactive,” Walls concluded.
Rosalyn Page travelled to AWS Re:Invent 2025 as a guest of AWS.