News
From Legacy to Leverage: How FPT Modernizes Systems for Enterprise AI Value
While much of the market is focused on bigger models, more durable value is being created by modernizing the legacy platforms that still power banking, insurance, manufacturing, and logistics. FPT is leaning into that space with a pragmatic AI-first approach, applying AI to decode decades of business logic, reduce modernization risk, and create clearer paths to cloud-native growth.
FPT is generating tangible value from artificial intelligence by modernizing large‑scale legacy systems worth billions of dollars, helping global enterprises optimize operations, control costs, and create new growth pathways in the digital era.
While much of the global technology conversation is focused on the race to build ever more powerful AI models, a quieter yet more sustainable opportunity is taking shape: the modernization of legacy IT systems. These systems continue to underpin critical industries worldwide - from banking and insurance to manufacturing and logistics. This is where FPT has chosen to compete, taking a pragmatic approach that applies AI to long‑standing, mission-critical challenges rather than pursuing technology for its own sake.
Legacy Code: The Hidden Backbone of the Global Economy
Behind the vast majority of global financial transactions lies a surprising reality: hundreds of billions of lines of code written decades ago. COBOL, a programming language developed in the late 1950s, is still estimated to account for 220–300 billion active lines of code worldwide. These systems have proven remarkably resilient, but their longevity has created new risks. As experienced engineers retire, maintenance becomes more difficult and expensive, while integration with modern platforms such as cloud and AI remains limited.
For enterprises, the challenge is not whether to replace these systems, but how to modernize them without disrupting the complex global transactions they support. Legacy platforms are not simply software assets; they embed decades of undocumented business logic. Transforming them is often compared to renovating an occupied building without access to its original blueprints - change must be incremental, carefully sequenced, and tightly managed to minimize risk.
Turning Six‑Decade‑Old Technology into Billion‑Dollar Opportunities
The rise of generative AI has fueled expectations that machines can automatically read, understand, and rewrite legacy systems. In practice, AI functions as a powerful accelerator rather than a replacement for human expertise. It supports tasks such as source‑code analysis, dependency mapping, and system reconstruction, but successful modernization still depends on deep domain knowledge and large‑scale execution capability.
This is why legacy modernization is not purely a technology exercise. It requires the integration of consulting, implementation, and long‑term operations - an end‑to‑end capability that FPT identified early as a strategic differentiator. Among its more than 1,100 global clients, a significant number continue to operate legacy systems, particularly in Japan. Instead of offering tools in isolation, FPT engages across the full transformation lifecycle, from assessing the current state and defining roadmaps to executing change and managing systems after deployment.
This model reflects a broader shift in FPT’s positioning: from a traditional IT service provider to a partner delivering high‑value services that are tightly integrated with customers’ core business operations.
FPT EVP & FPT Software CEO Pham Minh Tuan, highlighted how AI is transforming legacy system modernization at FPT Techday 2024
AI in Practice: From Millions of Lines of Code to Measurable Results
One illustration is FPT’s xMainframe solution, which applies AI to analyze and decode legacy systems. In a project for a major Japanese steel manufacturing group, FPT used AI to examine more than six million lines of code, reducing the effort required in the initial assessment phase, which is widely regarded as the most complex and costly in modernization initiatives, by approximately 30%.
To date, FPT reports involvement in the modernization of more than 300 systems, transforming over 200 million lines of code for more than 40 enterprises globally. These figures reflect the scale and depth required to succeed in this market.
The complexity of such efforts is evident across the industry. The New York Times began migrating away from COBOL in 2006 but halted the initiative three years later, only restarting in 2015 and completing the transition two years after that. Similarly, Commonwealth Bank of Australia spent five years and USD 750 million - twice the original budget - to rewrite its core system in a modern programming language.
These initiatives are measured in years, not months, and are closely linked to stable IT budgets. As a result, value is created through long‑term consulting and operational services rather than one‑off implementation projects.
A Large, Sustainable Market with High Barriers to Entry
Legacy IT modernization is emerging as a significant global market as organizations pursue digital transformation without abandoning mission‑critical systems. According to Mordor Intelligence, the market is projected to reach USD 29.39 billion in 2026 and expand to USD 66.21 billion by 2031, representing a compound annual growth rate of 17.64%.
Against this backdrop, FPT has significantly strengthened its legacy modernization capabilities through Flezi NEXT - an AI‑enabled portfolio designed to modernize mainframe, midrange, and complex legacy environments at enterprise scale. Built on more than 15 years of legacy modernization experience, Flezi NEXT brings together methodology, talent, and AI accelerators into a unified modernization framework. To date, FPT has delivered more than 100 successful modernization projects, analyzing and transforming over 300 million lines of legacy code across industries including BFSI, manufacturing, automotive, and healthcare.
Flezi NEXT, FPT's AI‑enabled portfolio designed to modernize mainframe, midrange, and complex legacy environments at enterprise scale
At the core of Flezi NEXT is Flezi EMT, FPT’s AI‑powered modernization platform, which embeds GenAI and agentic software engineering across the full lifecycle - from discovery and reverse engineering to rewriting, rebuilding, testing, and operations. Flezi EMT incorporates over 80 specialized AI agents and a legacy‑specific knowledge base accumulated over 15+ years, enabling enterprises to reduce project preparation time by up to 70%, lower reliance on scarce legacy experts by around 70%, boost developer productivity by 50%, and shorten overall modernization timelines by approximately 30%.
Flezi NEXT is underpinned by FPT’s “Winning 3A” approach to accelerating modernization: an Augmented Workforce that blends experienced legacy engineers with AI‑augmented delivery teams; Autonomous AI Agents embedded in accelerators such as code analysis, migration, and testing; and AI‑driven adaptation across methodology and delivery processes. This approach enables FPT to support both AI‑enabled rewriting and AI‑enabled rebuilding strategies - helping clients transition to open, cloud‑native, no‑vendor‑lock architectures while carefully managing risk.
The combination of platform, people, and industrialized execution is reflected in real‑world outcomes. Using Flezi NEXT and Flezi EMT, FPT has helped a leading Japanese steel manufacturer modernize a Fujitsu mainframe‑based MES system to cloud‑native Java and React, reducing integration costs by around 50% while improving scalability and performance. In another engagement, a large Asian insurance company rewrote a finance management system built on AS/400 and PL/I, cutting maintenance costs by 70% and doubling development speed. Similar agentic rebuilding programs for European industrial clients have delivered up to 80% faster release cycles and 50% lower operational costs.
FPT focuses on long‑term investment in people and digital workers to accelerate legacy modernization at large scale
Equally important is FPT’s long‑term investment in people and digital workers. Flezi NEXT is supported by a deep global talent pool of more than 3,500 COBOL, RPG, and PL/I engineers, 4,500+ cloud engineers, 1,500+ data and AI specialists, and 200+ industry SMEs. These teams work alongside AI‑powered digital workers - embedded across assessment, code generation, testing, deployment, and operations - to preserve institutional knowledge while continuously reskilling delivery teams for modern cloud‑native environments.
As a result, FPT’s AI‑driven growth in legacy modernization is not built on short‑lived technology trends, but on an execution‑driven model that combines domain expertise, scalable AI platforms, and sustained people development. In a market defined by complexity, risk, and longevity, Flezi NEXT positions FPT to help enterprises modernize with confidence - turning decades‑old systems into resilient foundations for future digital growth.
