SYSTEM.INIT(COBOL)

For years, rip and replace has dominated conversations about the mainframe. A smarter, more pragmatic reality is now taking hold. Mainframe systems can process millions of transactions in seconds, making them indispensable for core business functions – from database management to ERP and CRM – that demand absolute consistency and speed. Far from being obsolete, they are the quietly powerful bedrock of modern commerce, handling 90% of all credit card transactions and serving over 70% of global enterprises, according to Gartner. But as digital demands grow, businesses need their mainframes to keep up. Modernisation helps them move with more agility, support developers better, plug into cloud technologies and manage costs more smartly, which are all essential to stay competitive. This forces a new, more urgent question for today's IT leaders. As every penny of investment is scrutinised, the focus has shifted from debating the mainframe's future to a more practical challenge. Namely, how do we modernise this critical infrastructure and clearly demonstrate its value in a hybrid world? Practical modernisation without the Big Bang ## In conversations I have with IT leaders it's clear the idea of a risky 'big bang' migration is losing its appeal and for good technical reasons. More than 70% of Fortune 500 companies still rely on mainframes – according to Forbes – often built on decades of interwoven COBOL and RPG code and custom business logic. In this environment, where a single change can trigger a domino effect across critical programmes, attempting a wholesale lift-and-shift presents too many risks to be considered a viable strategy. Instead, a more practical, incremental modernisation is proving effective. The key is to connect mainframes with cloud environments to create a hybrid ecosystem that leverages the best of both worlds. A key technical pattern is to API-enable core mainframe applications, to allow them to participate seamlessly in this broader architecture. This allows teams to build new, cloud-native front ends that leverage the unmatched security and reliability of mainframe transaction processing on the back-end. There’s no denying this approach demands substantial investment, time and meticulous planning. But this approach to modernising without dismantling allows teams to innovate securely, reduce technical debt and ensure every change is targeted and aligned with clear organisational objectives. It’s an approach that allows teams to innovate without introducing unnecessary risk, systematically reduce technical debt and ensure that every single investment in modernisation is directly aligned with clear, measurable business objectives. How AI is changing the performance conversation ## AI is accelerating this modernisation shift by directly addressing the longstanding challenge of data gravity, and recognises that the smartest approach is to bring analysis to the data instead of forcing the data to move. Rather than undertaking the costly, complex and high-risk process of moving massive volumes of sensitive information to a separate platform, AI models can now run directly on the mainframe. This provides, for the first time, clear business insights from raw performance data, right at the source. A retail bank, for example, can analyse live transaction data to understand why a customer might abandon a purchase, all without that information ever leaving the secure mainframe environment. Building on this, generative AI is accelerating the modernisation process itself, offering powerful new ways to analyse legacy systems and streamline transformations. This technological shift completely reframes the investment conversation. Leaders can now justify mainframe spending not as a defensive capital expenditure to maintain the status quo, but as a growth-focused operational expenditure that directly underpins business success and a better customer experience. Total cost of ownership as a strategic tool ## And this is exactly where Total Cost of Ownership (TCO) becomes key. As AI reshapes how organisations extract value from the mainframe, IT leaders need a way to quantify that value clearly and credibly across the business. Modern TCO can’t stop at a basic comparison of hardware costs. In the AI era, the conversation needs to mature from TCO to Total Business Value. This means calculating the ROI of the platform by factoring in the strengths that drive real impact: built-in security that reduces breach risk, resilience that prevents costly outages, and the transactional performance that underpins core business operations. The 2025 BMC Mainframe Survey confirms the mainframe is here for the long haul, with a 97% long-term commitment rate. This reality brings a new question to the forefront, namely how do you manage it intelligently? The answer lies in a value-based TCO, which provides the framework to make the crucial "modernise vs migrate" decision for every application. For one financial services giant, this meant standardising how TCO was calculated across more than 4,500 applications. By creating a unified data model that captured this broader view of value, they gained the clarity needed for an honest, apples-to-apples comparison. This is what makes modern TCO a strategic aid in business as it provides one source of truth needed to turn a complex choice into a clear business decision. Future proofing ## By tapping into AI for sharper insights, looking at TCO through a value‑driven lens, and taking a hybrid path to modernisation, leaders can get far more out of their most important systems and stay central to today’s digital business landscape. But this is just the beginning. The future of the mainframe lies in its ability to evolve, and AI is the key accelerator. As we look ahead to an era defined by AI and even quantum computing, the platform's unique strengths mean more applications, not less, will be best-fit for the mainframe, delivering unmatched value for the years to come.

“Charles Babbage was the father of computers… COBOL and FORTRAN were the first programming languages.” They were fifth standard trivia to memorise and earn marks for a whole generation of millennials. As they go about their daily work, COBOL has made an unexpected return as the central theme in modern tech conversations. The 67-year-old programming language is considered as the hidden backbone of the digital economy. Be it banking transactions or power distribution or government infrastructure, billions of codes behind several vital functions have remained untouched for decades, thanks to the terrifying complexity of changing software systems built using COBOL decades ago. Now, GenAI promises to replace legacy software, rewrite codes and modernise systems at a speed and cost unimaginable till recently. A clean erasure of technical debt in weeks. But what is technical debt in the first place? It is the hidden cost of years of patchwork fixes, messy or outdated code, legacy systems, disconnected tools, and poor or missing documentation that accumulates over time. As it builds up, it becomes a drag on the organisation, slowing innovation, driving up costs, and making it increasingly difficult to adopt newer technologies like AI. The opportunity unlock ## Resolving this tech debt or legacy modernisation has been a core services line of India’s $280 billion IT services sector. AI has both a disruptive and innovative impact on legacy modernisation. Large players like Tata Consultancy Services, Infosys, HCLTech, TechMahindra and LTM have acknowledged revenue deflation in the segment. But they are also winning larger deals because of enhanced scope of work. “Legacy modernisation has moved from “lift and shift” to full-stack reinvention,” explained Phil Fersht, chief executive of HfS Research. “We are seeing more scaled programs, especially in banking, insurance, and telecom, where estates are complex and regulatory pressure is high. Deal sizes are expanding, not just because of migration, but because clients are bundling modernisation with cost takeout and business transformation.” Customer Relationship Management (CRM) software leader Salesforce said global C-suites are no longer viewing modernisation as a cost-absorption exercise. “Historically, technical debt was seen as a back-office burden; today, it is the primary friction point slowing the shift toward an autonomous future,” said Prakash Thekkatte, senior vice president - software engineering, Salesforce India. “You cannot layer transformative AI onto a foundation of fragmented silos. To lead in the agentic era, organisations must first solve the 'data debt' equation.” HDFC Bank began modernising its technology stack in 2020. Its internal Neev platform brings AI models, APIs, controls into one place and connects AI agents to both legacy and modern systems. Ramesh Lakshminarayanan, chief information officer (CIO) at HDFC Bank said that although AI coding assistants have made modernising legacy systems like COBOL easier, it cannot rely on code generation alone. “We have to make sure the final system is safe, well‑designed and meets all regulatory and security requirements,” he said. “Today, about 35-40% of our coding is already done with AI…that meets security and regulatory norms,” he said. Enterprise Resource Planning (ERP) software leader SAP said that modernisation has become a major catalyst of SAP’s AI pipeline. “...we are seeing investments set aside with a focus on reducing technical debt, move to standard AI enabled processes, workflows and archive historical data from system of records to system of intelligence with an objective to maintain a clean and lean digital core,” said Mukesh Kumar H, head-customer advisory, SAP India. He explained that AI has made the cost of carrying technical debt far more visible and customers realise that fragmented landscapes and inconsistent data directly limit their ability to scale. Salil Parekh, chief executive, Infosys explained that for a transportation sector client Hertz, “we helped with a legacy migration to bring 3 million lines of COBOL code to a modern microservices environment using AI foundation models. The cost was 60% lower, the timeline was 60% quicker,” he said in a post earnings call. “What we are seeing is a decisive move away from rehosting, lift‑and‑shift, and narrow code conversion to an end‑to‑end re‑imagination of systems,” Satish HC, executive vice president & chief delivery officer, Infosys, told ET. “AI is enabling a reliable understanding of complex legacy estates, accelerating timelines, reducing risk, and compressing time‑to‑value. This is unlocking large modernisation programs and enabling enterprises to systematically address deep technology debt at scale, making this a durable growth opportunity for IT services,” he added. However, these may just be some early signs. Jimit Arora, chief executive intelligence and advisory firm Everest Group said that clients are not jumping all-in yet and their appetite to touch mission critical systems remains low. “With regards to COBOL, because of growing compute and cloud costs, clients who are on mainframes are very happy with the level of cost and performance. So even if capabilities exist, clients need a real business reason to switch away from the mainframe environments,” he added.

Foundational alliance supports phased migration and hybrid architecture approaches for enterprises managing mainframe-dependent infrastructure TOKYO, April 28, 2026 /PRNewswire/ -- OpenLegacy Japan Co., Ltd. (Shibuya-ku, Tokyo; Country Manager: Masahiko Shimoyama) announced a strategic alliance with Deloitte Tohmatsu LLC, a member firm of the Deloitte Tohmatsu Group, to support the modernization of legacy mainframe environments across Japanese enterprises. The alliance combines OpenLegacy's patented modernization platform technology with Deloitte Tohmatsu's system consulting and implementation capabilities to deliver flexible, low-risk modernization services. Under the terms of the alliance, Deloitte Tohmatsu will be responsible for upstream system design, leveraging OpenLegacy Hub to automatically generate digital services that bi-directionally connect legacy mainframe, midrange, and database systems with modern cloud and microservices environments. This approach supports the strangler fig pattern for incremental modernization without requiring changes to core mainframe code, allowing organizations to migrate function by function while maintaining continuity of mission-critical operations. Background ## "Japanese enterprises are facing a convergence of pressures that makes the status quo increasingly untenable," said Masahiko Shimoyama, Country Manager of OpenLegacy Japan. "The biggest challenge in Japan's '2035 Cliff' is a critical shortage of skilled talent to bridge the gap between aging legacy systems and modern AI-driven technology. As the workforce shrinks, the loss of technical knowledge risks leaving outdated systems permanently unmanageable, leading to a fatal decline in global competitiveness." Japan's Ministry of Economy, Trade and Industry has warned that continued dependence on legacy technical debt could result in significant economic losses as this specialized workforce retires. For organizations in financial services and other regulated industries, the scale and complexity of full system replacement presents significant operational, financial, and business continuity risks. Demand has grown for modernization approaches that allow organizations to transition in stages, or to operate legacy and modern systems concurrently, rather than undertaking wholesale migration. Modernization Approaches Supported Under the Partnership ## The partnership is structured to support two primary modernization scenarios, which can be pursued independently or in parallel: Phased Migration - OpenLegacy Hub's automated services generation enables bi-directional integration between new serverless Java environments and existing legacy systems at the functional unit level, without hand-coded bridges or proprietary middleware layers. By connecting individual components via microservices, APIs, and Lambda functions, organizations can migrate incrementally, beginning with peripheral systems and progressing toward core functions. This approach is designed to reduce migration risk, shorten implementation timelines, enable innovation during modernization, ensure full operational consistency, and minimize disruption to existing data integration specifications. Customers have reduced time-to-market for new services by up to 10x compared to traditional migration projects. Hybrid Architecture (Coexistence) - In cases where organizations choose to retain mainframe infrastructure for core functions such as high-volume transaction processing, while migrating front-end or agility-dependent functions to cloud environments, the partnership supports concurrent operation of legacy and modern systems. Data integration between environments is managed through OpenLegacy's API layer, allowing each system to operate according to its strengths. This model also reduces mainframe workload over time, with corresponding reductions in operational and infrastructure costs - with customers reporting cost reductions of 60% or more compared to maintaining traditional ESB and middleware stacks. Deloitte Tohmatsu will further support full-stack migration scenarios using its proprietary innoWake™tool, which provides automated code conversion from COBOL to Java and covers applications, data, batch jobs, and online screens. Strategic Context ## Beyond near-term migration planning, the partnership addresses a structural barrier facing enterprises seeking to adopt AI and data-driven capabilities. Legacy systems that lack modern connectivity cannot readily expose data to AI pipelines or integrate with cloud-native tooling - an estimated 88% of enterprises report that core legacy data remains inaccessible to generative AI models due to proprietary system formats and the absence of standard interfaces. OpenLegacy Hub is designed to make legacy system data accessible to modern environments without requiring underlying system rewrites, delivering production-ready digital services in as little as hours and days. OpenLegacy's technology is deployed by a global roster of Fortune 500 clients across banking, insurance, retail, manufacturing, healthcare, government, and telecommunications. Reference engagements include Legal & General, which achieved 10x scalability improvements by adopting an AWS cloud-first strategy supported by OpenLegacy Hub - a deployment that also validated OpenLegacy's position as a strategic partner within the AWS Partner Network - and AIG, which reduced claims payment processing time from eight days to minutes after exposing core insurance logic as digital-ready APIs through the platform. Deloitte Tohmatsu's Application Modernization Studio Tokyo, established in April 2025 as the firm's fifth global modernization center, provides a hands-on environment where enterprise IT and DX leaders can evaluate modernization approaches prior to implementation. OpenLegacy Japan is the Japanese subsidiary of OpenLegacy, a software startup headquartered in New Jersey, USA. Leveraging OpenLegacy's patented technology, the company provides solutions tailored for the Japanese market that rapidly, securely, and efficiently generate APIs (Application Programming Interfaces) to seamlessly connect mainframe, midrange, and database systems - which run on enterprise core systems - with cloud and microservices-based operating environments.

Agentic AI tools are helping organisations overcome Cobol skills shortages and untangle legacy infrastructure, but successful modernisation still requires an expert in the loop to manage complexity Skills, cost and agility are the three main drivers for organisations considering agentic artificial intelligence (AI)-supported code modernisation, according to Michael Vincetic, Kyndryl’s practice leader for cloud, core enterprise and zCloud in Australia and New Zealand. The skills issue stems from a persistent shortage of people with mainframe – especially Cobol – expertise. Addressing this gap reduces the risks associated with supporting legacy applications. “That’s attracted a lot of media attention lately with things like the Anthropic announcement,” said Vincetic, referring to claims that Claude Code can automate much of the work needed to translate systems written in Cobol into modern languages. However, as Gartner distinguished vice-president analyst Manjunath Bhat pointed out: “There is very little value merely in porting code from one language to another without modernising the architecture and infrastructure. It defeats the purpose because we don’t benefit from the scalability and reliability benefits of modern architecture patterns.” “The other reason is that modernising applications using composable and modular architectures helps adopt proven software engineering practices such as independent testability and independent deployment. These practices reduce the blast radius of changes and therefore minimise associated risks.” According to Anthropic, Claude Code does much more than simply translate languages. The company claims the AI understands dependencies, preserves business logic while modernising to current frameworks, and generates both test units and modern documentation from legacy code. “Agentic AI is already playing a major role in code modernisation projects from the most mundane to the most complex,” said Bhat. “When it comes to mundane activities, think of auto-implementing the system using modern design patterns, creating cloud landing zones, auto-generating new code, as well as creating the tests and deployments needed to verify if the functionality works.” A related benefit of agentic AI is that it can provide a way to operate this mainframe-centric infrastructure using natural language instructions, Vincetic pointed out. As for cost, he said: “The unit cost and commercial models for running on traditional infrastructure have shifted dramatically with the advent of new technologies.” Previously, trying to understand the logic within and around legacy systems was often more expensive than rewriting them entirely. But, Vincetic added: “AI has flipped that script now, and provided the capability to understand some of that detailed logic and interdependency that’s built up over, in some cases, tens of years.” Bhat agreed: “AI becomes more useful when we use it for semantic conversion rather than just syntactic conversion, by mapping out the underlying data flows and therefore ensuring that the new system works as expected. “Think of AI-assisted complex activities as the ability to explain what the code does, which parts of a complex system should be modernised first, which parts are most risky, and what the interdependencies are – both architectural dependencies and inter-team dependencies. These activities are part of what we might consider ‘discovery’ in the planning process. The more insights we glean using AI at this stage, the more prepared and seamless the downstream aspects of modernisation will be.” If an organisation’s goal is to increase agility, that probably means moving at least parts of the wider system onto a hyperscaler, Vincetic suggested, allowing them to take advantage of evolving capabilities in areas such as data management and analytics. When an organisation runs a mix of legacy and contemporary infrastructure, there is a risk of “two-speed IT”, where the former operates in a rigid, waterfall manner and the latter in a dynamic, agile way. Running multiple operations capabilities usually incurs more costs and hampers speed to market, especially during digital transformation initiatives, he warned. Banking and government are two of the primary sectors focusing on modernisation. Banks are being driven by ongoing digitisation projects, while governments want to improve the quality of citizen services. Kyndryl’s 2025 State of mainframe modernisation survey found that among the 80% of organisations that shifted their mainframe modernisation strategies over the past year, 43% are modernising more on the mainframe, 34% are integrating more with the cloud, and just 16% are moving more applications off the mainframe. Notably, only one of the 500 respondents planned to move off the mainframe entirely. However, the predominant driver for modernising code is to eventually move it off the mainframe, Vincetic noted. Kyndryl’s survey found that 98% of respondents are moving some applications off the mainframe, migrating an average of 28% of their workloads to other platforms. Conversely, 56% are increasing their overall use of mainframes, in part by positioning them as the centrepiece of a hybrid environment. Whichever strategy is in play, Vincetic suggested there are three key elements to modernisation: modernising the infrastructure (for example, moving from a mainframe to a public cloud), modernising the operations capabilities (to cope with the characteristics of the new environment), and modernising the overarching operating model. Extracting value from modernisation requires capabilities such as code conversion, but these only address specific aspects of the project. The real trick is understanding the context and interdependencies within a complex ecosystem like a mainframe. Agentic AI can perform much of the code conversion, but more importantly, it can untangle legacy systems to determine business rules and other characteristics. What’s more, “it can do it in probably about half as much time … at a high degree of quality,” said Vincetic. A mainframe application typically involves a set of data flows and integrations built over many years, alongside strict controls around data and policy management. Consequently, “the expert in the loop is still very critical,” he added. These controls cover regulatory compliance, availability and the disaster recovery requirements set by the organisation. “If you really want to extract full value, you need to fundamentally re-architect by and large, which then brings AI to the table,” said Vincetic. Mainframes have survived for many years because they are highly secure, incredibly good at processing high-volume transactions, and highly available. “There are certain workloads that absolutely are best placed to reside on a mainframe for those intrinsic capabilities,” he explained. “But that’s always countered by pressures from the market to become more agile, to present more data, and to leverage digital channels.” Therefore, it is a balance between preserving the intrinsic characteristics that make the mainframe so effective, and providing citizens or customers with the modern digital capabilities they expect. Modernising a system carries the implied goal of achieving equal or better capabilities than before, and that requires human expertise. Modernisation isn’t an all-or-nothing endeavour. Addressing one workload at a time is a practical approach, but understanding the context of that workload and the overarching drivers for modernisation is absolutely critical for success, Vincetic concluded. While AI can assist, expert guidance remains essential.

Tax season in the U.S. served as another reminder of COBOL's critical role in government systems and the ongoing efforts to modernize the language. Fresh off another tax season, U.S. workers are again reminded that much of their government’s infrastructure still runs on COBOL-based systems that have supported tax processing for decades. While the legacy language may feel dated to some, it retains a critical place, meaning COBOL skills will remain in demand while modernization efforts continue. Consider recent government initiatives to modernize COBOL: The U.S. Department of Health and Human Services announced early 2026 it replaced a legacy COBOL-based payroll system with a secure, cloud-based solution. <a href="https://www.hhs.gov/press-room/hhs-replaces-legacy-payroll-system-improving-service-delivery.html" target="blank" rel="noopener noreferrer">Link</a> Conversely, the IRS spent about $1.5 billion in 2024 to modernize systems, only to pause in 2025 to reprioritize, according to a Government Accountability Office report. <a href="https://www.gao.gov/products/gao-25-107611" target="blank" rel="noopener noreferrer">Link</a> The Risks of All-or-Nothing Replacement ## No matter the type of mission-critical transactional processing that relies on COBOL, the language can be used with modern technologies—including hybrid cloud, newer languages, APIs and AI/ML. However, the challenge with COBOL applications is that they rarely exist in isolation, cautions Scot Nielsen, vice president of product management at Rocket Software. They are generally embedded across business processes and systems in ways that might not fully be realized. “These applications are often among the largest in the IT portfolio, sometimes running to millions of lines of code,” he says. “Approaching them as straightforward rewrite projects can underestimate the scale, complexity and potential for disruption involved.” With many applications in use today created close to 50 years ago, all-or-nothing modernization may look attractive. Rather, Nielsen recommends incremental, strategic updates that reduce risk and disruption to modernization efforts and daily business activities. “If the assumption is that these systems are a problem to be eliminated, that framing can shape decisions in ways that don’t always lead to the intended outcomes,” Nielsen notes. “A more effective approach is to begin with a clear understanding of the role these systems play in the business.” The solution, he recommends, is an incremental approach that focuses on addressing technical debt, updating systems and broadening expertise organization-wide: “The starting point should be recognizing what you have as an asset, not a liability.” How to Plan for Modernization ## A COBOL modernization plan should be treated as a multistep approach that spans across the tech stack and business setting. IBM notes reasons to modernize COBOL-based applications that include: Compliance and security: Legacy code might not adhere to current data privacy controls and policies, or include the latest security patches, opening it to vulnerabilities. Interoperability: Communicating with modern business applications can ensure innovation and business value. Maintainability: COBOL codebases can be bloated and outdated. Their tightly coupled components make debugging and upgrades difficult, causing teams to spend time finding and addressing issues. When considering modernizing COBOL-based applications, Nielsen encourages strategic planning in the following order: 1. Gain visibility: Organizations need to understand what they have, what the code does, how it connects to business processes and where dependencies exist. 2. Define the business goal: Whether it is faster feature delivery, regulatory compliance, improved integration or cost optimization, the objective should guide the approach. 3. Recognize skills and institutional knowledge: Understand who maintains these systems today, how that knowledge is distributed and how to build broader expertise across the organization. “It’s important to understand what you already have,” Nielsen says. “Before any modernization step, organizations need visibility into their COBOL estate, including what the code does, how it connects to business processes and where potential risks lie.” Only then should organizations move into technical execution, including architecture, tooling and whether to take an incremental or larger-scale approach. Bridging the COBOL Skills Gap with AI-Assisted Tooling ## When it comes to concerns around skills related to COBOL, Nielsen notes that understanding the language is often cited, where the reality is in understanding the scale and complexity of the applications on which it runs. “Understanding how the application supports the business and making changes confidently at that scale are where the real challenge lies,” he cautions. As a parallel challenge, engineers who may have been around when these long-ago applications were first created are retiring, resulting in loss of both COBOL syntax knowledge and how the business systems operate. Because of this, before beginning modernization efforts, Nielsen recommends first ensuring proper tooling and documentation. “Once the application becomes more accessible and better understood, developers from different backgrounds can contribute more effectively, and the perceived skills gap often begins to narrow,” he highlights. Nielsen encourages teams to lean into modern AI-assisted tooling to document and ingest COBOL codebases at scale. Beyond quickly understanding these applications, he notes the benefits to include accelerated onboarding of new team members and cross-team knowledge sharing, rather than having this information leave with retirees. “Bringing developer toolsets up to par with current best practices is an early and high-value step that lowers the barrier to entry for new developers,” he says. “The result is a single, unified engineering team — COBOL developers working alongside Java, Python or cloud-native colleagues using the same tools, workflows and delivery cadence.” The Path to Continuous Value Delivery ## Teams should view COBOL modernization as an incremental approach that is a continuous practice that delivers value along the way, rather than an all-in-one destination, Nielsen advises. Each step should improve value and reduce risk over time and make progress, even as priorities, experiences, or staffing evolve. “The IRS example is a cautionary tale decades in the making, and one that any organization managing large-scale COBOL systems should study carefully,” he says. “What looks like a 2024 pause is, in reality, the latest episode in a 25-year pattern of attempts to replace systems the organization may have never fully understood in the first place.” Links: * <a href="https://www.hhs.gov/press-room/hhs-replaces-legacy-payroll-system-improving-service-delivery.html" rel="noopener noreferrer">HHS Replaces Legacy COBOL Payroll System, Delivering Faster, More Reliable Services</a> * <a href="https://www.gao.gov/products/gao-25-107611" rel="noopener noreferrer">IRS Is Developing a New Modernization Framework </a>

The Y2K bug turned out to be a non-event on January 1, 2000. How did that happen? Carl and Richard bring together a number of stories from folks who were there, fixing the software and updating systems, so effectively that, ultimately, nothing much happened when the clocks rolled over. It was common practice with early software to only store two digits worth of year - back then, storage space was at a premium. For years, there had been warnings about fixing these problems, but by 1999, it was essential. These are the stories of how some folks did those fixes so effectively that when Jan 1 2000, came around, nothing bad happened. Listen to the episode: <a href='https://www.spreaker.com/episode/how-we-beat-the-y2k-bug--71580154' target='_blank'>https://www.spreaker.com/episode/how-we-beat-the-y2k-bug--71580154</a>

Software Engineer Unni Siva Sankar, who has nearly two decades of mainframe migration experience across major US financial institutions, on why human expertise remains irreplaceable in COBOL-to-cloud transformations Menlo Ventures’ “State of Generative AI in the Enterprise” report shows that 76% of AI use cases in enterprises are currently purchased rather than built in-house, compared to 53% in 2024. This shift reflects companies recognizing that building specialized AI capabilities internally often takes longer and costs more than acquiring external expertise. Yet for one of enterprise IT’s most complex challenges, mainframe modernization, ready-made AI solutions don’t exist. Financial institutions face a particularly challenging version of this problem: migrating decade-old mainframe applications to modern cloud platforms while processing millions of transactions daily without downtime. Unni Siva Sankar spent 19 years performing just such migrations at Nordstrom and Wells Fargo, delivering millions of dollars in annual savings by eliminating IBM licence fees while maintaining zero defects in regulatory compliance. His unusual expertise, fluency in both 1970s COBOL and modern Golang microservices, backed by AWS Solutions Architect and Azure certifications, makes him one of the rare professionals capable of bridging the skills gap that stalls most enterprise modernisation initiatives. We spoke with Sankar about why AI-powered code transformation tools still require human expertise, what makes tax season the most brutal test for any financial system, and how his work on building Kerala’s first private hydroelectric plant connects to reducing computing’s carbon footprint. Businesses’ spending on artificial intelligence upgrades is rising, but most projects to migrate legacy systems still fail before reaching production. When converting a 30-year-old COBOL credit processing system that handles billions of transactions, what is the most dangerous assumption companies make about using artificial intelligence tools to accelerate migration? I saw projects fail because teams trusted AI-generated transformations without understanding that the original COBOL wasn’t just processing transactions — it was encoding decades of institutional knowledge about fraud patterns, regulatory exceptions and modern cloud-native design. I’ve seen projects fail because teams trusted AI-generated conversion without understanding that the original COBOL wasn’t just processing transactions – encoding decades of institutional knowledge about fraud patterns, regulatory exceptions, and edge cases that nobody documented. You need a human who can read both the old code and the compliance requirements. Most developers avoid legacy technologics, preferring to work with new tools. In 2006, you took a different path by choosing to master COBOL, JCL, and Endevor – legacy mainframe technologies that most developers were abandoning. What did you see that others didn’t? I noticed that all large financial institutions were completely dependent on mainframe systems to perform critical operations: credit card processing, transaction settlements, reporting to regulators and the people who understood these systems were getting older. It was obvious that companies would eventually have to modernise, but they would face an obstacle because no one would be able to read the existing code well enough to migrate it safely. Studying COBOL, JCL, and Endevor meant recognising that the rarest skill in corporate IT would be the ability to translate between legacy and modern platforms. By 2020, when financial institutions finally decided to move to the cloud, this foundation made me one of the few people who could analyse the business logic of a mainframe system, determine what needed to be preserved and what was obsolete, and develop the right Golang replacement that would not lead to regulatory violations. It takes an institutional understanding of why these systems were built a certain way and how seemingly redundant code actually prevents errors that cost millions of dollars. Migrating Nordstrom’s credit system saved millions of dollars a year by eliminating IBM licence fees. But during the transition, the old mainframe had to continue processing every transaction flawlessly while you built its replacement. How do you maintain zero defects in a production system that you are actively decommissioning? You treat it like a passenger plane that needs an engine replacement during flight. I had to demonstrate that we could reproduce every piece of the mainframe’s business logic in Golang microservices without introducing differences in transaction processing. We completely redesigned the processing model from mainframe batch operations to event-driven microservices, while proving that we could maintain transaction integrity under peak loads. The AWS deployment required security and high availability configurations, as any downtime would cost more than the entire project budget. But here’s what complicated the task: I spent my nights debugging production issues in 30-year-old COBOL when mission-critical batch jobs ended in failure, and then spent my days designing the architecture for new cloud services. The outdated system could not be allowed to deteriorate – business operations depended on its ability to process millions of transactions daily without a single error. Most organisations cannot find people who can operate effectively in both worlds simultaneously, which is why so many modernization projects stall or fail catastrophically. At a major financial services company in the U.S. , you managed tax form generation systems during the 2021-2022 tax season. These applications generate forms for millions of customers in accordance with IRS deadlines. What happens when a critical batch job fails, and the forms must be ready the next day? You have about six hours to identify the root cause, implement a fix, ensure it doesn’t corrupt data, and get the forms back into production because missing deadlines for tax forms means regulatory sanctions and significant fines. Tax season is when you discover what your systems are truly capable of, because peak loads stress-test every component. As a Technology Lead, I was fully responsible for systems that could not fail. This role required deep technical knowledge to debug issues through code-level analysis, as well as leadership skills to coordinate global teams under tight deadlines. When you’re rushing to fix batch processing of tax data for three million customers with only hours left before the deadline, every decision has to be right the first time. In addition to managing crisis situations, I implemented automation to reduce manual work and incident frequency in a highly regulated environment where every code change requires thorough testing. EBCDIC encoding has been obsolete everywhere except mainframes since the 1980s, but it remains the standard for financial data in these systems. During your work at Infosys, you created your own ETL tool in Golang to automatically perform this conversion. Why couldn’t you just use existing data migration software? Standard ETL tools treat EBCDIC conversion as a simple character encoding translation, which is a fundamentally flawed understanding of the problem. EBCDIC represents decades of mainframe-specific data structures, packed decimal formats, and binary encodings that have no direct equivalents in modern systems. When transferring credit card data from a mainframe to cloud storage, you are translating data types, processing obsolete field definitions that made sense in 1985, and preserving numerical precision that is critical for financial calculations. Commercial tools either couldn’t accurately handle these conversions or required manual intervention, which led to errors. My Golang ETL pipeline was designed to understand these mainframe data structures, automatically convert EBCDIC files, verify results against known correct outputs, and deliver clean data, reducing what used to take weeks to hours of automated processing with an accuracy unattainable by manual methods. In 2018, you co-founded Mukkudam Electroenergy to build Kerala’s first private small hydropower plant in the Idukki region, an area known for landslides and geological instability. Given that your design specifically incorporated seismic and landslide resistance measures, what drove you to take on such an engineering challenge, and what’s the connection between solving problems in unstable terrain and your work in mainframe migration? Engineering challenges have more in common than one might expect. In both cases, you need to understand the existing constraints, whether it’s 30-year-old COBOL or the topology of the Idukki river valley, and develop solutions that work within the real-world limitations. When I identified the site’s potential and conducted the initial feasibility analysis, it was the same problem-solving approach I use for system migration: assessing what is possible given the constraints, and then executing despite scepticism. Securing funding from IREDA required persistence because my co-founders and I were proposing something unprecedented. I redesigned the pipeline system to gain an additional 20 metres of head, which directly increased the power generation capacity. Beyond the engineering work, we incorporated a social objective into the project – we supply 10,000 units of energy per month to the local government hospital in Adimali, providing a reliable power supply for medical operations. The Hindu, Economic Times, and Times of India covered the project because it demonstrated private renewable energy development serving both economic and community needs. Kerala’s mountainous terrain makes it prone to landslides and geological instability. How did you address the engineering challenges of building a hydropower plant in such conditions, and what broader applications could your approach have for renewable energy infrastructure in geologically sensitive regions? The Idukki region experiences significant monsoon rainfall and has a history of landslides, which meant we couldn’t simply follow standard hydropower construction practices. We conducted extensive geological surveys and incorporated multiple safety measures into the design: reinforced foundation structures, strategic placement of the intake and powerhouse to minimize terrain risk, and continuous monitoring systems that can detect ground movement or structural stress. The most critical innovation was our approach to pipeline routing by redesigning the path to gain that additional 20 metres of head, we simultaneously chose a route that followed more stable geological formations and reduced exposure to potential landslide zones. This design philosophy of letting natural constraints guide engineering solutions rather than fighting them has broader applications. Many regions with high renewable energy potential: mountainous areas for hydro and wind, seismically active zones for geothermal, face similar challenges. Our project demonstrates that with careful site analysis and adaptive engineering, these geographical challenges can be converted into design advantages. The monitoring systems we implemented could serve as a template for other decentralized energy projects in unstable terrain, particularly in developing regions where renewable infrastructure is needed but geological risks have deterred investment. In 2024, data centres consumed 4% of the total electricity in the United States, and this figure continues to grow as the adoption of artificial intelligence accelerates. You have experience in both enterprise IT modernisation and renewable energy. As someone with this combination of skills, how do you feel about reducing the carbon footprint of computing? Modernising outdated systems has direct environmental implications that most people overlook. IBM’s licensing encourages the use of old equipment well beyond its useful life. When we migrated Nordstrom’s credit system to Golang microservices on AWS, we not only saved on licensing fees but also moved to a more energy-efficient infrastructure per transaction. But the picture is complex because it depends on server load, data centre power sources, and whether the cloud provider uses renewable energy sources. AI workloads are extremely energy-intensive, which can offset efficiency gains from other improvements. I am exploring how to apply the same approach I used in IT modernization, understanding both existing infrastructure constraints and cleaner alternatives, to help organisations reduce the energy consumption of computing systems. This could mean optimizing algorithms to require fewer compute cycles, better workload scheduling to leverage renewable energy, or infrastructure design that minimizes cooling requirements. My dual background positions me to work on problems at the intersection of both fields.

NEW YORK, April 16, 2026 /PRNewswire/ -- Kyndryl (NYSE: KD), a leading provider of mission-critical enterprise technology services, today announced that it has been selected by the North Carolina Division of Motor Vehicles (NCDMV) for a contract to upgrade its core motor vehicle systems. The initiative aims to significantly reduce customer wait times and expand online access to essential DMV services statewide. NCDMV is undertaking a comprehensive effort to replace its COBOL-based systems with a modernized platform as a significant part of an overall digital transformation to better serve North Carolina residents. The enhanced NCDMV environment hosted on Microsoft Azure supports faster, more resilient operations and future enhancements, while helping to meet rising driver expectations for reliable and user-friendly services. "North Carolinians deserve a modern, efficient DMV that makes essential services easier and faster," said Paul Tine, Commissioner, NCDMV. "I'm excited for the future of our agency with this partnership. This was the best deal for taxpayers, and I'm confident NCMAX will deliver a significantly better experience for both our customers and our dedicated staff." "Departments of motor vehicles provide essential services that people rely on every day, and the technology behind those services must be reliable and built to evolve," said James Rutledge, President, Kyndryl U.S. "By applying our experience modernizing complex, mission-critical environments, we are helping NCDMV build a more resilient foundation that improves everyday experiences and supports better service delivery for residents today and long into the future." NCDMV's new environment will support faster in-office transactions, expanded online services and improved customer engagement. The new system is part of NCDMV's strategic plan to increase self-service options, such as mobile driver license applications and renewals, reduce customer wait times and accelerate staff onboarding. Kyndryl is also helping to build an NCDMV Mobile App and Digital Wallet for payments, renewals, appointments and alerts. NCDMV selected Kyndryl based on its proven experience modernizing departments of motor vehicles across the United States. In large-scale initiatives in Arizona and Virginia, Kyndryl has helped agencies improve system reliability, enhance security and deliver more efficient, customer-focused services. About Kyndryl ## Kyndryl (NYSE: KD) is a leading provider of mission-critical enterprise technology services, offering advisory, implementation and managed service capabilities to thousands of customers in more than 60 countries. As the world's largest IT infrastructure services provider, the Company designs, builds, manages and modernizes the complex information systems that the world depends on every day. For more information, visit www.kyndryl.com. Kyndryl Press Contact ## press@kyndryl.com

Most mainframe users who turn to AI for help migrating legacy code to alternative platforms are going to be very disappointed, according to analyst firm Gartner. “More than 70 percent of mainframe exit projects initiated in 2026 will fail to produce the intended benefits due to an overestimation of generative AI tooling capabilities,” states a paper the firm published last week titled “Too Big to Fail: Why Mainframe Exit Projects Are Likely to Fail in the Age of Generative AI.” Gartner also thinks that the market for AI-powered mainframe migrations is set to pop. “By 2030, 75 percent of vendors operating in the ‘mainframe exit’ market will either pivot their business models or cease to exist,” the firm advises. The main reason for Gartner’s pessimism is the role of the mainframe as home to mission-critical applications and decades’ worth of data. “For most large-scale enterprises, the sheer volume and interconnected complexity of this data make wholesale migration a physical and financial impossibility,” wrote Gartner’s Dennis Smith, Alessandro Galimberti, and Tobi Bet. The trio also acknowledge that mainframes are a significant source of technical debt and note that generative AI is very useful when helping organizations to detect and describe that debt. But the analysts find generative AI has “significant limitations when it comes to the automated conversion and migration of legacy code.” “It also does not account for the unique capabilities that the mainframe offers (e.g., ensuring that the same performance and throughput is achieved after the migration).” Gartner’s team thinks one reason vendors are suggesting AI for mainframe exit projects is “Aggressive investor demand for AI capabilities as the sole indicator of a vendor’s long-term health forcing vendors to deploy AI even where unnecessary.” That pressure meets users’ concerns about difficulties finding staff to operate mainframes, and technical debt. AI can sometimes feel like the answer. Gartner advises wariness due to what it describes as “The gap between the ‘marketing promise’ of generative AI and its actual capabilities in code transformation.” “The stakes of a miscalculation are immense,” the analysts wrote. “Poor decision making regarding migration is not merely a budgetary overage; it is a threat to business and operational continuity.” “Falling for “seemingly magical solution’ migration promises while ignoring a platform-smart approach (i.e., diligently evaluating your workloads and choosing the best platform for the relevant work) leads to massive technical debt and critical enterprise risk.” The paper offers advice on how mainframe users should plan to use their big iron in future and suggests most should continue to look for ways to improve their systems rather than making a move. “The drive to abandon the mainframe is diminishing,” the analysts wrote. “Customers are increasingly recognizing the near-impossibility of a mainframe exit at an acceptable cost and risk, leading them to give up on the long-held hope for a perfect tool to achieve this migration.” Gartner’s opinion will go down very well at IBM, which saw its stock price slide sharply after Anthropic touted the COBOL-conversion powers of its Claude Code tool, sparking yet another round of speculation that the mainframe might be on its last legs. Big Blue’s revenue – which is currently swelling due to unusually high mainframe sales – suggests its big iron has plenty of life in it yet. Indeed, Gartner’s paper ranks the mainframe as “still the leading platform for certain mission-critical applications, even with the ongoing drive toward cloud-native architectures.”

COBOL remains a critical language for mainframe systems, yet existing large language models (LLMs) struggle to generate and translate COBOL code correctly. This paper reports our experience in developing and evaluating domain-adapted LLMs for COBOL and mainframe software engineering. We introduce: 1. an automated data curation pipeline that combines compiler-guided validation with multi-stage similarity-based filtering to construct high-quality COBOL training data, and 2. COBOL-Coder, a COBOL-specialized LLM fine-tuned on the curated COBOL domain data. We evaluate COBOL-Coder on two tasks: code generation (on COBOLEval and COBOLCodeBench) and code translation (on COBOL-JavaTrans, our proposed benchmark for bidirectional COBOL-Java translation). In our experiments, COBOL-Coder achieves up to a 73.95 percent compilation success rate and 49.33 Pass-1 on COBOLEval, compared to 41.8 percent and 16.4 for GPT-4o, while most open-source baselines (e.g., CodeGemma, CodeLlama, StarCoder2) fail to produce compilable programs. For Java-to-COBOL translation, COBOL-Coder reaches 34.93 Pass-1, whereas general-purpose LLMs achieve near-zero scores. To assess the usability of LLM-generated code in real-world settings, we conduct a survey with experienced COBOL developers. Participants consistently report that COBOL-Coder exhibits stronger COBOL awareness, has more reliable program structure, and is better aligned with enterprise practices than general-purpose LLMs.