Five Pressures Reshaping Mainframe Operations

Mainframe operations are evolving rapidly as organizations balance AI adoption, hybrid IT, DevOps, cost optimization, and a shrinking talent pool. This article explores the five biggest trends reshaping enterprise mainframe environments—and how IT leaders can prepare.

The mainframe isn’t going anywhere, but the environment around it, and the expectations placed on the teams that run it, have changed faster than most operating models have.

Experienced engineers are retiring and taking decades of institutional knowledge with them. Leadership wants cost comparisons against cloud alternatives. DevOps pipelines now extend to Z whether ops teams are ready or not. Generative AI is changing everything. And in a hybrid environment, a failure that starts in a cloud service can end on your mainframe before anyone realizes what happened.

Our teams are seeing the following five pressures shaping the evolution of mainframe operations:  

1. The Knowledge Crisis is a Structural Risk, Not Just a Staffing Problem

The mainframe skills shortage has been discussed for years, but what’s happening now is different in kind, not just degree. The engineers who built these environments are retiring, and in most organizations, their knowledge hasn’t been captured anywhere. The result is a structural vulnerability that won’t fully reveal itself until something breaks.

At the same time, recruiting remains difficult, with few developers wanting to enter the mainframe space. Those who do are expensive and scarce. As a result, some organizations are forced to make architecture and operational decisions based not on what’s optimal, but on who they still have.

2. Cost Optimization is Now a Core Mainframe Priority

For a long time, mainframe costs were treated as a fixed line item: significant, but not something that needed active management. That’s changing. CFOs are asking for cost-per-transaction comparisons against cloud alternatives. Procurement teams are modeling Tailored Fit Pricing scenarios. Finance wants a clearer picture of MLC software licensing exposure. And as cloud costs have come under the same scrutiny, the conversation has become more rigorous on all sides.

Mainframe teams are now being asked to instrument workloads, model consumption, and optimize MSU usage in ways that weren’t historically part of the role. When the work is done with rigor, the numbers can be substantial — the State of Florida streamlined its software spend and identified $8.3M in savings by approaching licensing and consumption deliberately rather than as a fixed cost.

Rightsizing is a live conversation, not a one-time exercise. And, in some organizations, the pressure to justify mainframe economics is driving architecture decisions that are more about cost optics than operational soundness. Navigating that dynamic — providing honest cost visibility without ceding ground to comparisons that don’t account for total risk — is increasingly part of what mainframe operations leadership has to do.

3. DevOps Expectations Don’t Stop at the Z Partition

The shift to Git-based workflows, CI/CD, and automated testing is now extending to the mainframe, imposing an additional burden on Ops teams, who aren’t staffed to account for it.

Traditional mainframe change control wasn’t designed for sprint cadences. Development teams working in modern IDEs have no visibility into what’s happening in mainframe pipelines. And the cultural gap between distributed systems developers and mainframe engineers can be as much of a barrier as the tooling: different vocabularies, assumptions about risk, etc.

Tools like Zowe, IBM Wazi, and IDz are advancing the practical integration of mainframe into enterprise DevOps workflows. Newer patterns, like the cloud-connected mainframe,  are pushing this further, making Z a more native participant in the same pipelines, identity systems, and developer experiences as the rest of the enterprise. But adoption requires more than tooling, it requires change in how mainframe teams think about development velocity, testing automation, and cross-team collaboration.

Teams that bridge this gap effectively are finding they can move faster and reduce the risk of changes at the same time. Those that don’t, are increasingly running two separate development cultures inside one enterprise, with all the coordination overhead that creates.

4. AI is Changing the Equation

AI really is a gamechanger for mainframe operations. Aside from the AI capabilities now available on Z, generative AI is being used to automate the identification and prevention of potential failures before they happen, transform mainframe application management, and modernize mainframe code in a fraction of the time it took before AI. AI is also accelerating code analysis, documentation, incident investigation, capacity planning, and operational automation across enterprise mainframe environments. This evolution requires a new breed of mainframe talent to bridge the gap, with the ability to keep operations running smoothly while also carefully adopting the cutting-edge technology that could just become an organization’s biggest differentiator.

5. Hybrid Observability is Essential for Modern Mainframe Operations

A decade ago, mainframe monitoring required deep expertise in one platform, knowing your SMF records, your RMF data, your SYSLOG. That’s still necessary, but it’s no longer enough. The mainframe now sits inside a hybrid environment that includes cloud platforms, APIs, distributed systems, and data pipelines — and a failure anywhere in that chain can surface as a mainframe problem, or cause one.

The challenge is that most monitoring and observability practices were built for a simpler architecture. They weren’t designed to correlate an API timeout in a cloud service with a downstream batch delay on Z, or to surface the relationship between a distributed system anomaly and a mainframe performance degradation. Closing that gap requires investment in tooling, in integration between platforms, and in building teams that can reason across the full environment, not just their corner of it.

Frequently Asked Questions

How is AI changing mainframe operations?

AI helps organizations automate diagnostics, improve observability, accelerate code analysis, optimize performance, and reduce operational risk across hybrid IT environments.

Why are hybrid IT environments increasing mainframe complexity?

Modern enterprises rely on interconnected mainframe, cloud, API, and distributed systems. Monitoring performance across these environments requires integrated observability and operational expertise.