The mainframe knowledge gap is a modernization risk

Much of an enterprise’s most important knowledge is also the hardest to access. Decades of institutional expertise can be deeply embedded in long-running applications and carried by specialists with years of experience navigating complex systems.

Modernization initiatives depend on a detailed understanding of the systems already running the business. Yet gaining that understanding can be difficult when applications have accumulated years of dependencies and operational complexity, particularly when the expertise needed to interpret those environments is concentrated among a relatively small group of people. As experienced specialists leave the business or simply face more demand than they can absorb, the difficulty of understanding critical systems can become a constraint on both day-to-day operations and long-term transformation.

The issue is especially visible in complex mainframe environments, where diagnosing performance problems can require deep familiarity with system behavior and specialized platforms such as CICS. The information needed to understand an issue may also be spread across multiple tools, adding another layer of complexity to the investigation.

For technology leaders, reducing these barriers to understanding is a critical part of modernization. Organizations need ways to make specialized expertise and operational context easier to access while preserving the depth of knowledge required to operate critical systems safely.

From institutional knowledge to accessible intelligence

Generative and agentic AI create an opportunity to address this problem in a different way.

Much of the discussion around enterprise AI has focused on producing new content or improving employee productivity. In complex IT environments, however, one of AI’s more consequential uses may be helping people interpret and apply the knowledge an organization already possesses.

That means using natural-language interaction to help employees make sense of operational information in the context of how their own systems behave. AI can help employees understand unfamiliar system behavior and determine an appropriate response. Where appropriate governance is in place, agentic capabilities can extend that assistance into the workflows used to address an issue.

This creates an opportunity to bring institutional knowledge and complex operational context into the flow of work, including for employees who may lack years of experience with the underlying systems. Understanding an issue can become less dependent on knowing which specialist tool to open or which expert to call first.

Why knowledge access belongs in the modernization strategy

Modernization decisions depend on context that may never have been formally documented. The rationale behind an application’s behavior or architecture may be scattered across runbooks and ticket histories, while critical knowledge often remains with experienced employees.

That makes understanding the current state an important modernization task in its own right. Teams need sufficient context to understand the implications of a proposed change before moving forward. Giving them easier access to that context can reduce the investigative work required to make modernization decisions confidently.

What this looks like in a mainframe environment

The opportunity becomes more tangible when applied to a specific operational issue.

Consider a CICS performance issue. An operations team might need to determine whether a recent software release contributed to a backlog or understand why response times have deteriorated. Answering those questions traditionally requires teams to gather diagnostic information from multiple tools and rely on an experienced specialist to connect the signals and interpret what they mean.

Rocket EVA is one example of how AI can change that process. The agentic AI platform is designed to help organizations understand complex mainframe and hybrid IT environments by connecting employees with relevant operational context and specialized expertise through natural-language interaction.

EVA’s capabilities include deep CICS analysis. An employee can ask why transactions are slowing down or whether a recent release contributed to queuing. EVA can bring together the relevant performance information, interpret what that data indicates, and help identify likely causes. From there, it can explain the issue in accessible terms and help teams understand an appropriate next step.

That progression matters because it reduces the tooling and expertise barriers that can make core systems difficult to understand, giving a broader range of employees a clearer path through the diagnostic process.

Expanding the reach of scarce expertise

Of course, making these systems easier to understand does not eliminate the need for experienced professionals, particularly when decisions carry significant operational or business consequences.

AI can increase the leverage of those specialists. Experienced employees can spend less time manually assembling diagnostic context or answering routine questions and more time addressing difficult problems. Less-specialized employees, meanwhile, can investigate more issues independently and bring better context to conversations with subject-matter experts.

The result is greater reach for expertise that may already be in short supply. It can also make organizational knowledge easier to apply because employees have a more direct way to connect that expertise with the operational information generated by the systems themselves.

A different measure of enterprise AI value

As CIOs shape their AI and modernization strategies, they should consider how difficult their critical systems are to understand as well as how scarce the expertise surrounding them has become. Reducing those barriers can help teams operate long-running applications more effectively today while making it easier to determine how those systems should evolve.

For organizations carrying decades of complexity in their core systems, enterprise AI offers a way to make that existing technology more manageable and its accumulated knowledge more useful as modernization moves forward.

See how Rocket EVA can help teams investigate mainframe issues faster and make operational intelligence more accessible.