AI: The Application Modernization Accelerator?

A Modern Conundrum

The irreplaceable value of core business systems has been a source of concern to IT leaders for some time. While these systems continue to prop up huge organizations, they face difficult challenges in delivering a fresh chapter of innovation and capability that markets often demand.  

First, the systems are expensive to maintain - but would be even more expensive to overhaul. Also, they require specialist skills to continue to maintain, that are harder and harder to find—but which would require far greater resource and expertise to significantly change or replace. Decades of application evolution, increased technical debt, and undocumented business rules are huge learning challenges even for skilled technicians.  

Furthermore, the troubling concern facing the CIO is that the net gain of such a risky undertaking is precious more than a functionally equivalent, slightly newer IT system. That’s hardly a cast-iron justification.  

The difficulties associated with justifying, planning, and executing large-scale application modernization programs are well documented.  

AI: A Contemporary Catalyst for Change

Beneath the endless cacophony of hype and hyperbole, artificial intelligence (AI) has already assisted countless enterprises with vital initiates such as task automation, predictive analytics, resource optimization, and cloud modernization. With greater power and potential, generative AI could be poised to take things further still. And 73% of survey respondents agree that generative AI will have the “most significant impact on enterprise organizations," according to a June 2023 article by researchers Vanson Bourne.

This has resulted in the lofty ambitions of major application modernization projects being reconsidered, with AI and GenAI new ingredients. The IDC FutureScape report titled, "AI Everywhere" Will Impact Business Decisions at Every Level (November 2023), included as one of its top ten predictions, “GenAI Transforms Application Modernization IT Services: Increased utilization of AI in application modernization IT services can streamline efficiency, enhance services delivery speed, and bolster IT services margins."

Futurum Research recently echoed the same perspective regarding the opportunity, “As organizations seek to leverage AI for efficiency and enhanced business value, the convergence of AI adoption and mainframe modernization surfaces as a crucial trend defining the future of enterprise IT applications and infrastructure.”

Making the Case

It is worth considering the power of GenAI in some typical modernization use cases, to understand its potential. To do so, let’s revisit our technical requirements from the previous blog.  

Complex Application Understanding and Modernization Planning

Applications earmarked for modernization are frequently highly complex, large-scale monolithic entities, housed in typically mainframe or midrange data centers. Specialist skilled staff of those systems have usually dwindled, and expertise lost. As such, the first and fundamental modernization use case—no matter what the technical strategy—is to attempt to understand the existing application in sufficient detail to truly understand the right options for change. Accelerated understanding and forensic insights, no matter what the legacy code or data format, is vital. Whether AI is infused at the analyst stage or (see below) thereafter, a further layer of innovation offered by sophisticated software solutions is vital in this fundamental first step. It is the door opener to any modernization strategy, from the eradication of technical debt to full-blown reengineering strategies.  

Augmented and Accelerated Change

Converting a so-called legacy application to a more modern incarnation can be tricky and risky. Previous generation "conversion tools" were of mixed value at best. AI-powered application analysis is a welcome string to the bow of forensic assessment and change automation, that can tackle all manner of existing codebase of long-standing applications, and then by leveraging machine learning algorithms, cluster analysis and programming language models, it can automatically generate refactored or optimized code, as and when required. More sophisticated variants will factor in elements such as language choice (a key consideration of any modernization project is to ensure the future state is flexible and not ‘locked’ to a proprietary format), user experience improvements, and the disaggregation of monolithic structures into more maintainable microservice-based components, more suited to a modern, often cloud-centric future state. This saves developers time, addresses technical debt, and reduces the chances of error.

Business Rule Extraction

Unsurprisingly, the same conceptual framework as above can and should be exploited in the pursuit of the true nuggets of gold buried in those core application sets—the business rules. These represent and reflect the core intellectual property that has been the nucleus of value delivered by IT for so many years. Isolating, mapping, and then extracting those business rules, again using AI-centric learning, pattern matching and modelling techniques, offers a fast-path mechanism to modernizing the highest-value elements of the core business systems, accelerating time to delivery of the most important changes.  

Additional Considerations

As we imagine our modernization project unfolding, additional technical requirements will come into play requiring the intervention of innovative, AI-based technology.

User Interface (UI) Modernization

As a company evolves, so does its design. Legacy applications often have outdated, inefficient user interfaces that may not meet modern design standards or user expectations. Generative AI can assist in automatically generating new UI elements, styles, and layouts to improve the user experience without requiring a manual redesign effort. The technology also ensures that the current best practices are followed, taking a load off designers’ shoulders.

Test Cases and Testing

By automating aspects of the testing workflow, generative AI assists in identifying potential issues, ensuring the reliability and robustness of the refactored code. Smarter AI tooling will propose changes alongside supporting test cases and—ideally—automatically-generated unit tests, to help quickly validate the changes. This streamlining of the testing phase of a modernization project contributes significantly to a more efficient delivery cycle, which traditionally would be consumed by expensive testing cycles (and resources) that can last up to 40% of the entire project.  

Conclusion

Modernization programs to support major business initiatives have recently become more viable, with huge advances in platform choice and supporting deployment software. However, they still carry huge risks and are highly complex technical challenges.  

Innovative technologies, now infused with the power of AI, offer fresh advantages in the vital role of ongoing application modernization. Furthermore, GenAI may also help justify broader modernization investments: TechTarget reported in October 2023 that, "66% of … tech investments would be easier to justify if they supported a GenAI initiative."

The aforementioned Futurum research brief set to be released next week summed it up very well. “The integration of AI is a central driver for the modernization of mainframe operations, enabling organizations to improve operational efficiency, foster innovation, and leverage insights from complex legacy data sources. Intelligent tools and technologies powered by AI—such as the AveriSource Platform™️—simplify the application modernization journey through automated code discovery and analysis, identification of areas for optimization, and an accelerated pathway to cloud-native architectures.”

Learn More

To experience how the AveriSource Platform's AI-driven modernization technology can help accelerate your transformation, book a demo with the AveriSource team of experts here.