In an era where mainframes continue to be the backbone of global enterprises, the challenges of maintaining, optimizing, and modernizing outdated legacy systems have become paramount. These systems, often burdened with technical debt and fragile from years of quick fixes, are not only costly to maintain but also hinder business agility. 

The scenario is further complicated by the impending retirement of a significant portion of the skilled workforce that has the knowledge to maintain these systems, as well as the often-inadequate technical documentation that makes knowledge transfer, mostly through person-to-person, expensive and inefficient.

At the same time, demand for mainframe modernization is growing, driven by the need for streamlined applications that enhance agility and mitigate continuity risks. However, these projects are notoriously complex, with a 74% failure rate among organizations attempting them.¹ Recent advances in technology, partnerships with Cloud Service Providers, usage of hybrid cloud environments, as well as incremental modernization strategies, offer more viable and less risky pathways to modernization.

Despite this, COBOL modernization continues to pose significant challenges. Decades of business logic encoded in COBOL make it difficult to extract, document, and translate to modern languages. Automated code conversion often yields Java that is correct but hard to maintain and extend, failing to utilize modern programming practices. Moreover, traditional one-size-fits-all tools don't fully address the nuances of different legacy systems, necessitating extensive manual effort in debugging, testing, and refactoring.

There's a growing recognition of AI's potential to address these challenges, offering promising solutions for automated code conversion, retro-documentation, and testing processes in mainframe modernization.

Can Artificial Intelligence and Generative AI address the challenges?

Generative AI (GenAI) represents a leap beyond traditional mainframe technology and AI applications. It's not just transforming business and societal problem-solving with new, human-like content; it also has tremendous potential for accelerating mainframe modernization. GenAI's deep understanding of legacy code semantics as well as heuristic interpretation of enterprise engineering standards allow for more effective capture of business logic and intent, enabling precise code transformation and knowledge encapsulation.

Several consulting firms, IT services, and Cloud Service Providers are exploring AI applications for code refactoring, generating visuals to demystify complex systems, and improving data migration. These innovations aim to bridge legacy systems with modern technologies, though many are still under development.

How are Capco and Wipro FullStride Cloud leveraging GenAI capabilities to accelerate mainframe maintenance and modernization?

Capco and Wipro FullStride Cloud have adopted a ground-breaking approach to tackling our customers' most pressing challenges with innovation and pragmatism. This approach has been developed through Capco’s extensive management consulting and industry experience together with Wipro FullStride Cloud’s industry leading cloud and infrastructure expertise. At the heart of our vision is the synergy of knowledge management (one of the most important business critical and continuation risks for our clients in all industry domains), existing COBOL code optimization, and the transformation of legacy COBOL applications to sustainable, modern systems. Note that we are not advocating mainframe exit using GenAI: Rather, by harnessing GenAI for a more systematic and controlled modernization process, we simplify legacy applications and prepare them for a future where flexibility, performance, and maintainability are paramount. 

Capco and Wipro FullStride Cloud’s proprietary set of accelerators creates virtual replicas of legacy applications by analysing millions of lines of legacy code. Feeding GenAI extracts of these virtual replicas allows it to accelerate modernization outcomes across knowledge management and code optimization, application modernization and testing, and application maintenance and support. 

Note that we do not promote one Large Language Model (LLM) over another. Generative AI models are becoming more sophisticated at an unprecedented pace, and many of the proprietary as well as open-source models perform exceptionally well for our stated purposes. 

Knowledge Management and Code Optimization

• For static content, we develop domain-specific knowledge portals that are tailored for specific types of users and usage scenarios. This eliminates a ton of technical heavy reports that typically gets created by current market tools which often goes unutilized. 
  
• Moreover, our ‘Ask AI Anything’ chatbot capability empowers users to extract insightful and meaningful answers from GenAI for complex technical queries, including business logic and business rules extraction, co-pilot for COBOL, or provide Intelligent system refactoring recommendations towards simplifying and optimizing existing legacy code.
  
• Our code analysis feature can also be used to conduct deep, semantic analysis to uncover security vulnerabilities patterns or performance bottlenecks in the source code. Unlike traditional tools, GenAI can understand the context of code, making it better at spotting complex vulnerabilities and suggesting more nuanced solutions
  

Application Modernization and Testing

• Wipro FullStride Cloud and Capco’s extensive knowledge management capability is ideal for reverse-engineering legacy source code to user stories and acceptance criteria and then building functional test cases on these criteria. This, together with having detailed knowledge on how data is persisted in the legacy system allows us to create synthetic data.
  
• In addition, we employ classical Machine Learning (ML) techniques to design sustainable, isolated service stacks to improve target system performance and reduce costs, and ensuring the modernized system is robust and maintainable. 
  
• Moreover, leveraging GenAI, we ensure that the final code is fully compliant with enterprise engineering quality and security standards, automate the creation of target system documentation as well as Unit Tests.
  

Application Maintenance and Support

• Our solution is very useful in the Incident Management domain: GenAI is used for Root Cause Analysis and pinpointing underlying problems in the source code, and then suggesting solutions based on the resolution of past incidents and remedies or by generating new solutions through AI’s contextual understanding of the application's codebase. As an extension of our Knowledge Management capability, we can automatically update the knowledge base by documenting such incidents, and the corresponding Root Cause Analysis and resolution.
  
• Our code scans can also turn into predictive mode based on trends and anomalies it finds in the code it is scanning and comparing it to the Root Cause Analysis of past incidents. These incident management opportunities can automatically be documented in the tool of choice thereby logging, tracking and audit is automatic without any manual intervention, while workflow approvals are directed to the right authority.
  
• We are also leveraging GenAI to improve developer productivity in application development and maintenance tasks by identifying opportunities for improving on existing issues in the source code or even best coding practices for ease of maintenance. For example, a large scope of analysing specific patterns in application code base and applying remediation for similar patterns is a good use case which we are exploring through a Proof of Concept for a customer use case.
• At a broader level, Capco and Wipro FullStride Cloud implement GenAI to manage complex dependencies in legacy systems, ensuring that updates or migrations do not break functionality and reducing integration issues in the process. Finally, it is important to acknowledge that integrating our tool into your DevOps pipeline allows for dynamic updates of knowledge, ensuring that system documentation remains current with the production code.
  

Conclusion

References

¹ 74% Of Organizations Fail to Complete Legacy System Modernization Projects, New Report From Advanced Reveals | Business Wire