In 2026, the role of the mainframe developer will undergo a profound transformation. Once seen as the guardians of foundational computing systems, today’s mainframe professionals are at the heart of digital innovation, bridging the gap between tradition and the future.
Mainframe teams are leaner than ever, requiring each developer to wear multiple hats. The days of siloed expertise are fading; now, developers must understand not only mainframe technologies but also cloud platforms, AI integration, and modern DevOps practices. Systems are more interconnected, with mainframes serving as the backbone for mission-critical workloads in banking, healthcare, and government. This interconnectedness means developers must collaborate across disciplines, ensuring seamless data flow and security between platforms.
The pace of change demands ongoing learning. Hands-on training, apprenticeships, and upskilling in areas like AI, cybersecurity, and cloud integration are now standard. While some companies are investing in cross-skilling and training to keep teams agile and future-proof, most are not, leaving developers to adapt quickly and learn new tools and methodologies on their own, while simultaneously maintaining these mission-critical systems, known for their stability and reliability. Developers are expected to adapt quickly, mastering new tools and methodologies while maintaining the reliability and stability mainframes are known for.
There’s More than One Kind of Mainframer
Izzi Software’s Chief Architect Chris Craddock – a veteran of BMC, CA, HPE, and other major technology companies – explains that there are actually two main groups of mainframe software developers. “The one that I come from is the group that builds system software. We focus on building the operating systems (or extensions) and commercial software products that add operational and business value, e.g. monitoring, automation, utilities etc.
“The second group are the folks who write business applications. They’re writing CICS transactions, IMS transactions, batch housekeeping programs, accounting things, and, to the extent that it’s still necessary, things like system exits and utilities. And then there’s this tiny slice of the Venn diagram between the two sides who are typically called systems programmers who need to cover aspects of both. So there’s system stuff on one side and business stuff on the other side. On the business side of things, people are spending their day writing and debugging COBOL programs that run as transaction programs. It’s really important to understand this distinction because the future is going to be very different for both groups as we move into 2026.”
Mainframe Engineers Matter
Mainframes are no longer just about stability: they’re surrounded by cloud and other platforms, and have to become more comprehensively integrated for business innovation and reduced operational costs. Pundits everywhere are talking about how AI will create new levels of functionality and intelligent automation without a programmer in sight. The good news for mainframers is that AI has not been trained on mainframes and they are different enough from the other platforms that AI will have a delayed impact. The bad news is there will be headcount pressure and mainframe engineers will be required to cross skill and learn to leverage their strengths, while integrating with new technologies that drive business value.
Despite predictions of obsolescence, mainframes are still central to many large enterprises, but they’re also complex with an aging user base. A new generation of talent entering the field faces challenges applying their own expertise to an unfamiliar platform. Our challenge is to help them to rapidly learn how to adapt to the mainframe and keep it at the forefront of a rapidly evolving IT landscape.