Cinq questions essentielles que les dirigeants devraient se poser avant d’entreprendre la modernisation de l’ordinateur central

By Hassan Zamat

Mainframe modernization has long been perceived as a complex, multi-year, multi-function undertaking. But the reality is that mainframe modernization has become surprisingly agile. Forward-thinking organizations are embracing an iterative modernization model, transforming mainframe applications in manageable increments.

Rather than pausing critical operations, enterprises leverage each modernization milestone — and its attendant ROI — as a catalyst for subsequent initiatives. Findings from our third annual Mainframe Modernization Survey Report show that this pragmatic approach is supported by flexible strategies, impressive financial returns, and the adoption of advanced capabilities such as agentic AI and generative AI.

I'd suggest that leaders preparing to initiate or accelerate their modernization journey ask themselves the following five questions. The answers will help guide you toward a pragmatic, and ultimately successful, mainframe modernization strategy.

1. Is your strategy truly flexible?

Organizations achieving significant business results through mainframe modernization consistently adopt strategies that are built for change. That means rigid, five-year modernization roadmaps are being replaced by shorter-term plans that are more adaptable. Even these short-term plans are subject to revision: our data reveals that 80% of leaders have changed their mainframe strategies within the past year.

This flexibility comes in direct response to evolving business needs and external pressures. The rapid advancement of AI technologies has heightened expectations across IT, with business stakeholders demanding increased data accessibility, enhanced legacy application functionality, and seamless AI integration. IT teams must therefore be prepared to pivot in response to shifting priorities and opportunities.

This ability to pivot relies on organizational culture as much as it does on technology. Not every team can — or wants to — turn on a dime. Leaders should critically evaluate both their vision for mainframe modernization and their organizational culture, and ask if they’re prepared for real-time adaptation.

2. Are you maximizing the value of AI?

AI has emerged as a transformative force in mainframe environments, driving operational efficiency, code modernization, and overall business value. The integration of generative AI, in particular, enables IT teams to automate code mapping, documentation and refactoring when needed.

One of our customers faced a familiar quandary: They were facing a critical skills gap due to the retirement of seasoned COBOL developers. Through the adoption of AI solutions, they were able to:

• Use generative AI to convert COBOL code to Java and generate comprehensive documentation
• Apply agentic AI to map dependencies and sequence cross-team activities
• Leverage large language models (LLMs) to distill application business logic into agile user stories
• Implement reinforcement learning to optimize developer workflows

The outcome was a 60% reduction in code analysis and documentation time, improved code quality and maintainability, and diminished reliance on scarce legacy skill sets.

Moreover, AI is expanding the ways organizations can use their data on the mainframe. While migrating mainframe data to the cloud was once the only option for AI enablement — a less viable path for highly regulated sectors — modern connectors and agents now allow organizations to integrate AI directly with on-premises data. In the near future, organizations will be able to run LLMs natively on mainframes, further increasing data control and flexibility.

3. Do you have the necessary skill sets?

Talent shortages remain one of the most significant challenges in mainframe modernization. Some 74% of companies in our survey say they rely on external partners to address capability gaps.

Successful modernization initiatives demand expertise not only in legacy and contemporary programming languages, but also in security, cloud integration, and AI operations. Then there's the ability to manage AI itself. Fortunately, this skill is attractive to existing and prospective IT employees, as it helps them develop in-demand capabilities.

Leaders can proactively invest in workforce development through targeted training, certification programs, and ongoing skills assessment.
 

4. Are you optimizing your ROI?

Mainframe modernization is delivering substantial returns. For the second year in a row, we’ve seen that the average ROIs from mainframe modernization projects are in the triple digits.

Effective strategies prioritize putting the right workload on the right platform, as determined by factors such as performance, cost, and regulatory requirements. Remember that Incremental, strategic steps deliver the best value. You don’t need — or want — an all-or-nothing strategy. An effective strategy aligns with business goals, preserves data sovereignty, and importantly, evolves as needed.

By resisting all-or-nothing approaches and focusing on continuous improvement, organizations can achieve superior agility, resilience and innovation.

5. Should you be running newer technologies on your mainframe?

Mainframes are evolving rapidly, and their integration with newer technologies — from agentic AI to DevSecOps — offers substantial competitive advantages.

We recently worked with a global retailer that wanted to optimize its inventory in real time without moving sensitive data. Their supply data was stored on mainframes, and their retail consumption data was stored in the cloud. By using agentic AI to bridge the two systems, this customer achieved just-in-time insights while preserving data sovereignty. Shelves stayed stocked. Warehouses stayed lean. And customers got what they needed, when they needed it.

Conclusion

Valued for their unparalleled reliability, security, and performance, mainframes continue to support mission-critical workloads. As the technological landscape shifts, mainframe modernization is essential to maintaining enterprise agility and unlocking new sources of value — particularly in the age of AI. Leaders who ask the right questions and foster a culture of flexibility and continuous improvement will be best positioned to realize the full potential of their mainframes and their modernization initiatives.

Hassan Zamat is Global Practice Leader, Core Enterprise and zCloud, for Kyndryl.