IBM Utilizes Generative AI to Modernize Mainframe Cobol Applications

IBM, a global technology leader, is taking strides in modernizing legacy Cobol applications with the power of generative artificial intelligence (AI). In a groundbreaking move, the company is working on a cutting-edge software development tool, the **watsonx Code Assistant for Z**, which leverages AI to facilitate the transformation of Cobol code to Java, breathing new life into mainframe-based applications.

The upcoming **watsonx Code Assistant for Z** is poised to revolutionize the way organizations approach the modernization of their Cobol-based systems. This innovative tool is underpinned by IBM’s proprietary **watsonx.ai code model**, boasting an impressive 20 billion parameters. It aims to empower developers with the capability to seamlessly translate Cobol code to Java, specifically tailored for IBM Z systems.

A bridge to modernization 

IBM’s ambitious venture comes in response to a critical demand in the industry. The *IBM Institute for Business Value* conducted research that highlighted the significance of mainframe-based applications in the strategies of IT executives. A comprehensive survey involving 200 IT executives in North America revealed that an overwhelming 68% of respondents deemed mainframe systems as central to their hybrid cloud strategy.

The research also shed light on a pertinent issue faced by organizations: while nearly 70% of surveyed executives acknowledged the need to modernize their mainframe-based applications, the task posed considerable challenges. The complexity of the task often discouraged the complete overhaul of application estates, and the study found that organizations were 12 times more likely to leverage existing mainframe assets instead.

Generative AI paves  the path for efficient modernization

IBM’s generative AI-driven approach serves as a beacon of hope for organizations aiming to update and modernize their applications efficiently. The primary advantage lies in developers’ ability to rapidly assess, update, validate, and test code, streamlining the modernization process and allowing them to concentrate on high-impact tasks. This approach not only accelerates the modernization of large applications but also facilitates the transformation of Cobol business services into meticulously architected, high-quality Java code, incrementally.

Industry leaders embrace generative AI for cobol modernization

Financial technology company Broadridge has enthusiastically embraced the potential of generative AI to rejuvenate legacy Cobol code. Roger Burkhardt, the CTO for capital markets and AI at Broadridge Financial, expressed his optimism about collaborating with IBM. He highlighted the productivity gains achievable through generative AI interfaces and the opportunity to revolutionize capital markets solutions.

The alliance between Broadridge and IBM marks a pivotal moment, as industry leaders look to harness the capabilities of generative AI for modernization endeavors. Burkhardt revealed that the response from clients to their generative AI investments has been overwhelmingly positive, fueling their interest in further utilizing IBM’s **watsonx Code Assistant for Z** to address an even broader array of platforms.

IBM’s strategic approach involves integrating the upcoming **watsonx Code Assistant for Z** with their **Application Discovery and Delivery Intelligence (ADDI)** inventory and analysis tool. The process commences with the utilization of ADDI, followed by refactoring business services in Cobol. Subsequently, the transformation of Cobol code to Java code with optimized design takes place, culminating in a meticulous validation process that includes automated testing capabilities.

Driving real progress through generative AI

Kareem Yusuf, the senior vice-president for product management and growth at IBM Software, expressed the company’s enthusiasm for the far-reaching impact of their generative AI capabilities. He underlined the targeted and optimized approach adopted by IBM for the development of the **watsonx Code Assistant for Z**, emphasizing its potential to rapidly and accurately convert code tailored for IBM Z systems.

Yusuf underscored the tool’s role in accelerating time to market while expanding the skills pool. This approach aligns with IBM’s commitment to enhancing applications, introducing new capabilities, and preserving the performance, resiliency, and security inherent in IBM Z systems.

The watsonx Code Assistant for Z is set to make its debut at the highly anticipated *IBM TechXchange* technical learning event in Las Vegas from September 11 to 13, 2023. This unveiling promises to showcase the tool’s capabilities and potential, offering attendees a firsthand experience of how IBM’s generative AI can revolutionize the world of legacy code modernization.

IBM’s ambitious endeavor to bridge the gap between legacy Cobol and modern Java applications has the potential to reshape industries. With generative AI at the helm, organizations can embrace a more efficient, streamlined, and impactful approach to modernization, ushering in a new era of application development and transformation.