Sustainable AI – Building on System Modernization

Sustainable AI in mainframe modernization

In our whitepaper “Sustainable AI for Enterprise Transformation, Innovation and Growth,” we have shown that the development of Sustainable AI is becoming an indispensable technology for managing "material" risks and sustainable operations in large organizations. It shows how AI, especially Generative AI, can be integrated with existing management systems to improve prediction, optimization and planning across multiple business functions. It enables data-driven decision-making and streamlines complex sustainability initiatives across departments and value chains.

An implementation based on Generative AI platform cores and flexible agents enables seamless access and integration of existing information silos, unstructured corporate documents, and human communications. Such platforms can also help deliver relevant process information to management in the language and level of detail they need to quickly develop actionable solutions. Without AI-based platform integration, management would remain blind to the potential of many complex and sustainability-focused opportunities.

At the same time, the cost of maintaining aging mainframe systems is rising dramatically. Finding and retaining programmers capable of fixing security flaws or adapting to new compliance regulations in decades-old code is an increasingly difficult task. As a result, organizations face increasing risks from their own operations as external, market, and environmental risks grow. Such operations are neither sustainable nor scalable.

Fortunately, the current wave of AI development, especially Generative AI, provides not only an incentive to modernize systems, but also the tools necessary to catch up in a much more cost-effective way than before. As explained below, legacy code can now be more effectively translated for cloud-based applications. But such necessary system modernization is only the first step. Value can be created when AI agents begin to help effectively transform existing applications for the potential of cloud environments and prepare them for integration into innovative business models.

Industries that rely heavily on legacy systems-such as healthcare, financial services, and the public sector-also face the highest demands for effective contributions to sustainable outcomes. By planning their systems modernization as part of a transformation to sustainable AI solutions, they can not only achieve significant cost reductions and resilience improvements, but also begin to innovate operations and entire value chains.

The growing market for mainframe modernization

The market for mainframe modernization is experiencing significant growth. According to Fujitsu’s latest research, the mainframe modernization market is projected to grow to $15.4 billion by 2027, with an annual growth rate of 5%. Organizations are willing to invest significantly $3.9 billion to address the limitations of their legacy systems. Despite budget constraints, the potential cost savings and risk mitigation associated with modernization make it a worthwhile investment for many organizations.

Key industries driving this market include banking, discrete manufacturing, government, and insurance.

Challenges of maintaining legacy mainframe systems

Organizations relying on mainframes face a multitude of interconnected challenges. The high cost of maintaining aging infrastructure directly impacts profitability. Mainframes often hinder business agility, making it difficult to integrate with modern systems and scale operations to meet evolving market demands. This lack of agility also complicates compliance with regulatory requirements.

Furthermore, the speed and sophistication of cybersecurity threats pose a continuous risk. Data breaches, impacting user information or financial transactions, can lead to significant financial losses, reputational damage, and legal repercussions. At the heart of these challenges lies the scarcity of specialized expertise. Finding talents with the skills to understand and update these complex systems is increasingly difficult, as many experienced professionals are retiring, and the market is not producing enough new talent.

The benefits of Sustainable AI mainframe modernization

Given these challenges, AI emerges as a crucial enabler of modernization. The benefits are multifaceted:

• Connected business: Modern applications can share data within the whole organization and easily access third party systems.

• Operational Efficiency: AI-driven automation streamlines code refactoring, testing, and maintenance, reducing time and effort significantly.

• Error Reduction: AI enhances accuracy in code translation, testing, and operational tasks, minimizing human error and improving reliability.

• Improved Agility: AI tools facilitate quicker adaptation to changing business needs, enabling streamlined modernization processes and seamless integration with modern systems and providers.

• Cost Management: Automation and predictive maintenance powered by AI optimize resource utilization and prevent downtime, leading to substantial cost reductions.

• Enhanced Development: AI tools provide invaluable support to developers through code assistance, improved documentation, more efficient testing, and streamlined troubleshooting.

• Sustainability Potentials: Modern AI platforms support the integration and management of “dual materiality” innovations for sustainable growth.

Beyond these technical benefits, a shift in mindset is equally crucial. Fujitsu’s consultants, for example, emphasize the adoption of DevOps practices, fostering collaboration and agile methodologies to accelerate software delivery and improve overall efficiency.

Migrating away from mainframe and legacy systems unlocks numerous benefits, enabling new applications to become integral parts of Sustainable AI systems. In these systems, modern applications can seamlessly share data and interact to generate new value, leading to increased profitability and a significantly reduced environmental footprint. This translates to lower greenhouse gas emissions, decreased CO2 emissions, reduced electricity consumption, and less waste. This vision of sustainable, efficient, and profitable technology is what drives modern enterprises.

Modernization using Generative AI

Modernizing mainframe applications has two core elements: the application source code and the underlying infrastructure. The source code is crucial for seamless integration into modern systems and easy updates. A simple line-by-line translation of COBOL to Java or C# would obscure the application's business logic and workflows, rendering it less manageable.

Generative AI tools such as IBM Watsonx Code Assistant for Mainframe, will analyze, interpret, and rewrite COBOL, written based on different design principles, into object oriented code suitable for modern architectures. IBM’s watsonx.ai LLM has the knowledge of hundreds of coding languages and can generate highly optimized code. Furthermore, integration with tools like Application Discovery and Delivery Intelligence streamlines the entire legacy mainframe modernization process, providing a comprehensive inventory and analysis capability.

Applications at the centre of the enterprise

Similarly, Amazon Bedrock can convert COBOL code into documentation and translate it into Python, Java and other modern languages. This approach opens the door to leveraging Bedrock's extensive toolset. The modern code becomes the source to build Generative AI agents that utilize high-performing foundation models from leading AI companies like AI21 Labs and Cohere. Techniques like fine-tuning and Retrieval Augmented Generation (RAG) allow for the customization of these modernized applications and their associated data.

This approach fundamentally transforms mainframe applications, moving them from obscurity within legacy systems to a central position in modern architectures, where they can leverage the latest technologies to generate significant value.

Cloud for sustainable infrastructure modernization

Modernizing legacy systems requires addressing both application source code and infrastructure. Once legacy applications are transformed, they can fully utilize AI tools to deliver new value. Sustainability can become an important objective for developing AI-driven solutions to boost efficiency, offering services ranging from data center optimization to carbon emission and energy consumption reduction.

With legacy processes now fully integrated, tools like Alibaba Cloud's Energy Expert leverage AI to process vast datasets, analyzing supply chains and providing real-time insights and actionable recommendations. This is particularly beneficial for companies like Unilever seeking to optimize their supply chains.

These modernized processes can be integrated into platforms like SAP's Sustainability Control Tower, a solution that consolidates ESG data across the entire organization. Similarly, Oracle Cloud's sustainability management platform incorporates analytics, AI, and IoT capabilities.

Furthermore, cloud providers such as Microsoft Azure, Google Cloud, and AWS offer carbon footprint tools that provide a comprehensive view of an organization's carbon emissions, broken down by service, region, and time period. These tools consider factors like regional energy mix, hardware efficiency, and server utilization rates, suggesting optimizations such as utilizing renewable energy sources, workload optimization, and the removal of unused resources.

Fujitsu's role in sustainable AI modernization

Fujitsu is at the forefront of using AI solutions for mainframe modernization. Fujitsu's PROGRESSION offers automated tool conversion from COBOL to Java or C#, a capability further enhanced by generative AI for code development and refactoring. AI provides real-time insights into system status and optimization opportunities.

Fujitsu's extensive partnerships with AWS, Microsoft, and IBM are vital to delivering comprehensive, modern, and sustainable solutions. These collaborations facilitate the integration of mainframes into hybrid cloud environments, leveraging the scalability and flexibility of the cloud. Cloud computing offers a cost-effective approach to hosting applications. In fact, collaborations with Microsoft and AWS have yielded solutions that reduce the Total Cost of Ownership (TCO) by 50%.

Savings migrating from mainframe to the cloud.

Different strategies can be used to migrate applications to the cloud. Each strategy offers distinct benefits, influencing the overall Total Cost of Ownership (TCO) savings.

Retain: This approach involves keeping applications on their current platform, using AI to optimize and manage them for efficiency.

Rehost: This strategy involves a "lift and shift" approach, moving applications and data from mainframes to a cloud platform with minimal changes.

Replatform: This strategy involves moving applications and data to a different target technology platform (e.g. from on-premises to cloud) while maintaining the application architecture.

Repurchase: This strategy involves replacing existing applications with standard Commercial Off-the-Shelf packages or Software as a Service (SaaS) solutions.

Rebuild: This strategy involves building a new cloud-native application from scratch, leveraging the latest tools and frameworks.

Rearchitect: This strategy involves a significant architectural redesign to leverage multi-cloud environments.

Refactor: This strategy involves AI to redesign and optimize existing applications.

Public sector success stories

As mentioned above, the public sector faces some of the highest demands for system modernization and sustainable operations. Fujitsu specializes in highly complex modernization projects that can help transform such organizations. The Virginia Department of Human Resources Management, for example, achieved $15 million in annual operating cost reductions. Similarly, the Washington State Department of Licensing realized a $1 million annual reduction in TCO and avoided up to $500,000 in mainframe upgrade costs. Their systems are now ready for further integration and support of sustainable management transformation.

Conclusion

Organizations with legacy mainframe applications are reaching a tipping point for systems modernization. AI, especially Generative AI, can now support strategic cloud migration and help transforming operations for more efficient, sustainable, and value-added results. Since most organizations that rely heavily on legacy mainframe systems - such as healthcare, financial services, and the public sector - also face the highest demands for sustainable outcomes, their AI-based transformation can gain the most by focusing on the implementing Sustainable AI systems during systems modernization.