Modernizing IT with AI: Quick, Cost-Effective, High-Quality

A significant impediment to agility lies at the core of many large organizations: the technical debt embedded in legacy IT systems. These systems, often developed decades ago, underpin critical business functions across various industries. Notably, approximately 70% of the software utilized by Fortune 500 companies is over 20 years old.¹ Historically, modernizing these outdated systems has been viewed as an "IT problem," with business leaders deferring action due to concerns about prohibitive costs, extended timelines, potential disruptions, and uncertain returns on investment.

Modernization projects can be expensive, time-consuming, and disruptive, leading to a preference for maintaining existing systems that "still work." However, modernizing IT systems has become a critical business priority as technology becomes increasingly central to value creation and competitive advantage. Advancements in technology—including generative AI (gen AI), cloud computing, and robotics—offer organizations unprecedented opportunities to reimagine their IT foundations. Modern technology infrastructures are essential for capturing value in today's digital landscape, where technology drives approximately 71% of successful business transformations.² Additionally, the impending retirement of programmers who maintain these legacy systems intensifies the urgency for modernization.³

‍The Transformative Role of Generative AI in IT Modernization

Generative AI is reshaping the cost-benefit dynamics of modernizing legacy IT systems. What once seemed prohibitively expensive and time-consuming is now achievable at a fraction of the cost and effort. For example, modernizing a transaction processing system for a leading financial institution—which would have cost over $100 million three years ago—can now be accomplished for less than half that amount using Gen AI.⁴

AI's capabilities in automating manual tasks can accelerate IT modernization timelines by 40% to 50% and reduce costs by up to 40%.⁵ Beyond cost savings, AI delivers higher quality outputs and enables businesses to track the monetary impact of technical debt on profit and loss outcomes, often accounting for up to 50% of total investment spend.1
‍

A Strategic Approach to Modernization with Generative AI

However, the value of generative AI is not solely in its technology but in its strategic application. Effective modernization requires focusing on the following strategies:
‍

1. Improving Business Outcomes: Modernization must go beyond merely updating old systems.
   ‍
2. Empowering Autonomous Gen AI Agents: Companies can automate end-to-end processes with human oversight by building and training a network of gen AI agents.
   ‍
3. Scaling Value Across the Organization: The actual value of AI-driven modernization lies in its scalability.
   ‍
4. Leveraging Fully Integrated Platforms with Retrieval-Augmented Generation (RAG): Modernizing legacy IT systems is most impactful when organizations adopt a fully integrated platform strategy incorporating advanced generative AI methodologies.

How Gen AI Agents Can Improve Both Code and Business Outcomes 

Technology modernization transforms existing applications to leverage modern technologies, frameworks, and architectures. This process involves rewriting applications in user-friendly programming languages, transitioning to modern frameworks, restructuring systems for modularity, and migrating applications to cost-efficient cloud environments. Generative AI (Gen AI) agents offer a transformative approach to achieving these goals by focusing on three critical areas: improving business outcomes, enabling autonomous Gen AI agents, and scaling value. 

‍

1. Improve Business Outcomes 

Historically, many organizations have adopted a "convert-and-go" approach, feeding legacy code into Gen AI tools to translate it into modern programming languages. However, this approach risks merely migrating technical debt into a new system—a trap reminiscent of early "lift-and-shift" cloud migrations, where organizations moved legacy applications to the cloud without addressing fundamental inefficiencies. 

Modernization efforts should prioritize improving processes and systems to drive meaningful business value rather than maximizing the number of lines of code converted. Properly applied, Gen AI helps translate complex legacy elements—such as code, observability data, and logs—into plain language descriptions in minutes. 

This clarity enables collaboration between engineers and business experts to: 

• Understand existing system functionalities. 
• Identify areas that create value for the business. 
• Discard unnecessary components and focus on essential modernization. 

For instance, a technology services company successfully leveraged Gen AI to reverse-engineer legacy documentation and generate clear guidance for system updates. The company streamlined its processes by involving technical and business stakeholders and achieved greater alignment on modernization priorities.  

Another example is that IBM utilizes Gen AI to generate related code artifacts, including Infrastructure as Code (IaC) templates and security configurations, facilitating a more holistic modernization process. This method ensures that modernization efforts are aligned with business objectives, leading to more efficient and effective outcomes. 

‍

2. Enable Autonomous Gen AI Agents 

The deployment of autonomous Gen AI agents represents a significant advancement in software development. These agents can perform specialized tasks such as data analysis, integration orchestration, and test case execution with minimal human intervention. By collaborating within a multi-agent framework, they can manage complex processes end-to-end, enhancing efficiency and reducing errors. 

By integrating AI agents into software development, Salesforce's Agentforce platform exemplifies this approach. This allows for deep language understanding and decision-making capabilities. This integration enables the automation of intricate workflows, accelerating development cycles and improving code quality. 

For example: 

• Data Mapping Agents: Analyze and document data structures, ensuring compliance and quality assurance. 
• Security Design Agents: Address threat analysis and information security. 
• Testing Agents: Continuously refine outputs and remediate errors identified during QA. 

Feedback loops enhance agent collaboration, allowing them to review and refine one another's work. Human oversight ensures alignment with business goals, and role-based IDs enable managers to trace and address issues efficiently. 

Real-world success stories highlight the impact of this approach: 

• A bank used Gen AI agents to modernize 20,000 lines of mainframe code, cutting the estimated migration time by 40%. 
• A global insurer improved modernization efficiency and testing by 50%, accelerating coding tasks by over 50% by employing reverse engineering and autonomous agents to automate workflows

‍

3. Focus on Scaling Value 

The transformative potential of Gen AI lies in scaling its impact across the enterprise. Many companies mistakenly focus on evaluating tools rather than developing comprehensive capabilities. Technology leaders should establish a central Gen AI capability of two primary components to scale effectively. In a Factory Model: A "factory" consists of teams that develop and manage Gen AI agents for specific end-to-end processes. These teams standardize tools and approaches, including monitoring, traceability, and access to large language models. Companies should initially develop five to ten factories to establish maturity before scaling further. In a more traditional  

Platform Model: A Gen AI platform like AWS or Microsoft Copilot Studio provides reusable services and capabilities to factories. Key components include user interfaces and APIs to integrate Gen AI services with enterprise tools like Jira or ServiceNow, Dynamics 365, or SAP. Additionally, they provide supporting services like data import or agent orchestration. Lastly, you can have a library of prebuilt Gen AI agents to execute specific tasks. This platform should be overseen by a dedicated team that tracks KPIs, such as feature utilization and service performance. By combining the factory and platform models, organizations can ensure that Gen AI's benefits are scalable, consistent, and tied to measurable business outcomes. 

Deloitte highlights the importance of developing agentic AI solutions that automate software development, sales, and marketing functions. By investing in such scalable AI capabilities, organizations can address numerous challenges simultaneously, thereby enhancing overall productivity and innovation. 

‍

4. Leverage Fully Integrated Platforms with RAG and Agentic RAG 

A fully integrated platform strategy amplifies the benefits of Gen AI by incorporating Retrieval-Augmented Generation (RAG) and Agentic RAG methodologies. These techniques combine generative AI's creativity with real-time data retrieval and decision-making capabilities, leveraging data the organization already has. Through RAG, organizations can begin by integrating AI with knowledge bases and legacy systems, and RAG ensures outputs are accurate, contextually relevant, and actionable. For example, RAG can automate system mapping by retrieving and contextualizing historical documentation, policies, and financial and operations data. Agentic RAG takes this further by allowing autonomous agents empowered by RAG to independently retrieve, analyze, and act on data while escalating unresolved issues to human managers. This allows for real-time problem-solving and decision-making across integrated platforms. 

For instance, Google's Vertex AI platform allows organizations to build and deploy machine learning models that integrate with existing systems, facilitating seamless data retrieval and processing. This integration enables AI agents to perform tasks such as system mapping and dependency analysis more effectively, accelerating modernization initiatives. Using an integrated platform with RAG capabilities, organizations can synchronize efforts across IT and business functions, enhancing efficiency and reducing duplication of efforts. This approach is efficient in industries with extensive legacy infrastructure, such as finance and insurance, where precision and compliance are critical. 

A New Era for IT Modernization

While generative AI is still in its preliminary stages, its potential to transform IT modernization is undeniable. By accelerating timelines, reducing costs, and enhancing output quality, gen AI is making previously unviable modernization efforts both feasible and strategically advantageous. As technology becomes integral to value creation, business leaders must proactively address technical debt and embrace AI-driven modernization. The stakes are high, but the rewards—improved business performance, agility, and competitiveness—make the effort worthwhile. Organizations that seize this opportunity to modernize their IT systems will be better positioned to thrive in an increasingly digital future.

Drive Modernization with ASCEND ^SM

As businesses face increasing competitive pressures, generative AI offers a powerful means to reduce technical debt, drive innovation, and align IT systems with long-term strategic goals. However, the actual value of modernization lies not in the tools alone but in how companies orchestrate them to create transformative outcomes. We address this challenge through ASCEND - a structured approach designed to help organizations navigate the complexities of modernization while delivering measurable, lasting results. 

ASCEND provides a diagnostic methodology that progresses through four analytical stages. The framework is built on six interconnected steps designed to assess and tackle modernization challenges systematically: 

‍

The ASCEND Framework focuses not only on technology but also on aligning modernization efforts with broader organizational objectives. Combining advanced AI methodologies with a structured, systematic approach delivers value in several ways. Reducing costs means automating repetitive tasks and addressing inefficiencies, which reduces operational overhead. Accelerating Timelines: Streamlined workflows and AI-powered solutions shorten development cycles. Driving Innovation: ASCEND empowers organizations to focus on strategic initiatives by freeing up resources and enabling faster iterations. Lastly, it enhances collaboration by allowing clear, actionable roadmaps to bridge the gap between technical teams and business stakeholders. 

The result is an approach that integrates innovative AI with a deep understanding of organizational needs. This enables businesses to modernize their IT systems while achieving measurable and sustainable outcomes. Through ASCEND, organizations can unlock the full potential of generative AI and position themselves for long-term success in an increasingly digital world. 

Next Steps: Moving Toward a Multiagent Orchestration Model

For companies ready to embrace a multiagent orchestration model powered by generative AI, the path forward requires a strategic rethinking of how technology initiatives are planned, executed, and scaled. By focusing on key areas of improvement, organizations can effectively modernize their systems, reduce technical debt, and unlock new value.

1. Rethink Large-Scale Technology Proposals

Traditionally, large IT modernization projects have required significant personnel and years of effort to deliver results. This legacy approach should be challenged at every turn. Any technology proposal that involves lengthy timelines and resource-intensive processes should be met with skepticism.

Review existing programs and proposals to evaluate how generative AI can shorten timelines and lower costs. However, only assume that some projects claiming to use generative AI will deliver value. Many initiatives may only incorporate AI in limited or superficial ways, falling short of their transformative potential. Be thorough in assessing whether proposed solutions genuinely leverage the full capabilities of generative AI to achieve meaningful results.

2. Tackle the Biggest Challenges First

Generative AI is uniquely suited to solving complex, high-cost problems traditionally considered too resource-intensive to address. Instead of spreading efforts across small-scale initiatives, which yield small-scale outcomes, organizations should focus on their most significant and complex technology challenges.

These might include:

• Multiyear modernization projects tied to critical business functions.
• Systems with significant technical debt that hinder innovation.
• Projects with costs in the hundreds of millions of dollars.

Revisit past plans deemed too expensive or time-consuming and reevaluate them, considering what generative AI can accomplish. The potential to radically alter the cost-benefit analysis of these efforts is one of the most compelling advantages of AI-driven orchestration.

3. Tie Modernization Efforts to Explicit Value

A successful generative AI initiative must be grounded in a detailed and dynamic business plan that explicitly ties outcomes to value. Too often, business plans focus narrowly on technology goals, ignoring the broader operating model and financial implications.

A robust plan should:

• Define the value at stake, both in total and increments, that can be achieved along the way.
• Consider all costs—ongoing infrastructure costs, one-time modernization expenses, and potential operational savings.
• Establish explicit activities and milestones to capture this value, emphasizing the break-even timeline.

Equally important is the discipline to revisit this plan regularly to ensure it remains aligned with business objectives. With such vigilance, plans avoid devolving into mere code delivery projects, losing focus on the broader value they were meant to achieve.

4. Prepare for Organizational Implications

Scaling a multiagent generative AI model will have significant implications for talent, technology, and operating models. These changes require thoughtful, proactive planning:

• Talent Strategy: Companies must invest in reskilling programs to prepare employees for roles in an AI-enhanced environment. This includes training on AI tools and fostering a culture that embraces collaboration between humans and AI agents.
• Operating Model Adaptation: As generative AI becomes integral to operations, companies must rethink workflows, governance structures, and decision-making processes.
• Financial Planning: AI-driven modernization will shift the balance between operating expenditures (OpEx) and capital expenditures (CapEx), requiring innovative budgeting and financial oversight approaches.

These are not solely IT concerns; they are strategic issues requiring attention from CEOs and the board. Companies that address these implications head-on will be better positioned to scale their AI capabilities and realize the full benefits of their investments.

The Opportunity Ahead

Companies are only beginning to tap into the transformative potential of generative AI for modernizing legacy technology and cutting down technical debt. Success in this space hinges on focusing efforts where they matter most: orchestrating generative AI agents to address meaningful business technology challenges.

By taking a disciplined and value-driven approach, organizations can go beyond incremental improvements to unlock significant innovation and competitive advantage. With generative AI as a strategic enabler, companies can reshape their technology estates into engines of efficiency and growth.