Carrying Your Black-Boxed Core Systems Into the Future, Without Stopping Them — Launch of AI-Powered Legacy Modernization and Migration Support

Carry your black-boxed core systems into the future—without stopping them. End-to-end support spans LLM-driven code analysis through AI-assisted phased migration and refactoring.

Amatsukaze Inc. has launched its AI-powered legacy system modernization and migration support service, helping organizations use generative AI to "understand" and "safely rebuild" legacy systems that have become black boxes over years of operation. The service provides end-to-end support spanning LLM-driven code analysis and documentation, phased migration and refactoring with AI coding agents, and modernization strategy grounded in cost-effectiveness — answering the deep-rooted challenge of organizations that want to rebuild but where no one fully grasps the whole picture anymore.

The Current Landscape: Even Past the "2025 Digital Cliff," Legacy Assets Remain Untouched ​

The "2025 Digital Cliff," which Japan's Ministry of Economy, Trade and Industry (METI) warned of in its 2018 DX Report, has come and gone. Yet what lay beyond it was not a scene of triumph. Surveys report that roughly 60% of companies still carry aging core systems, and many of them fear a negative impact on their business. The essence of the problem the DX Report identified — the "black-boxing" of systems — persists and has, if anything, only deepened as the generation that built these systems retires.

The heart of the problem is not the code itself, but the absence of people who can understand the code. Core systems patched together over decades contain countless business rules, exception handling, and workarounds embedded without ever being documented. Because this tacit knowledge cannot be grasped, migration is risky, its cost cannot be estimated, and the recurring decision has been to "leave it untouched and mothballed." Soaring maintenance costs, a graying technical workforce, and a vanishing pool of maintainers are management risks that can no longer be deferred.

Generative AI is now triggering a tectonic shift in this stalemate. A symbolic moment came in February 2026, when Anthropic published its "Code Modernization Playbook." It showed that AI coding agents can analyze legacy code such as COBOL, mapping dependencies across thousands of lines, documenting business workflows, and surfacing risks — tasks that once took months — in a fraction of the time, and it drew a strong reaction from the industry. At the same time, experts including IBM pushed back, arguing that "translating code is not all there is to modernization" and that "decades of optimization tightly coupled with the hardware cannot simply be ported." This debate illuminates a core reality: AI dramatically accelerates analysis and understanding, but the success of a migration still hinges on human and infrastructure-operations judgment.

Why Legacy Modernization, and Why Now ​

There are three main reasons to start now.

• The "cost of understanding" has flipped: For years, understanding legacy code cost more than rebuilding it—the primary barrier to modernization efforts. LLM-driven analysis inverts this equation, creating an environment where even assets long deemed "untouchable" can finally be tackled.
• The interest on technical debt keeps compounding: Neglected legacy assets continue accruing interest through soaring maintenance costs, security risks, and slow adaptation to business change. The longer you defer, the more both migration cost and business risk grow.
• The talent cliff has become real: As the generation that designed and maintained core systems retires, tacit knowledge is being lost rapidly. The value of using AI to extract and visualize knowledge from the codebase — before person-dependent know-how disappears entirely — is rising.

That said, even though AI has accelerated analysis, "deploy a tool and migration completes itself" is not how it works. Validating the extracted logic, choosing the migration approach, planning a cutover that does not stop the business, and ensuring stable post-migration operations all demand a high level of expertise and a real operational perspective. This is precisely where a partner who understands both AI and infrastructure adds value.

What Outcomes You Can Expect ​

Pursuing legacy modernization delivers these benefits:

• Eliminating the black box: We regenerate lost documentation from the codebase, making specifications, dependencies, and business logic visible. Understanding and risk assessment of the entire asset advance, and the prerequisites for decision-making fall into place.
• Minimizing migration risk: Through phased migration with automated test generation, you can run the old and new systems in parallel and switch over gradually. If something goes wrong, the scope is small and reversible.
• Structurally improving maintenance cost: By breaking free from dependence on old languages and platforms, you eliminate soaring maintenance costs and reliance on increasingly scarce specialists, putting your IT cost structure on a healthy footing.
• Agility for business change: Migrating to a modern stack unlocks cloud scalability, data utilization, and rapid addition of new features — turning your systems from a drag on the business into a driving force.

How We Make It Happen ​

Rather than aiming for a wholesale rewrite from the outset, Amatsukaze adopts an "understand first," low-risk approach as its core principle, supporting you from analysis and strategy through phased migration.

• AI-driven code analysis and documentation: Using LLMs, we extract and visualize specifications, dependencies, and business logic from existing systems whose documentation has been lost. This accelerates the understanding and risk assessment of black-boxed assets and lays a solid foundation for migration.
• Modernization strategy: We evaluate options such as rewrite, replatform, and refactor on a cost-effectiveness basis, and define priorities and a business-continuity-first migration roadmap — making clear where and how to begin.
• AI-assisted phased migration and refactoring: Leveraging AI coding agents, we drive migration from old frameworks and languages to a modern stack. Through phased refactoring accompanied by automated test generation, we rebuild safely while verifying behavioral equivalence.
• Cloud migration and high-load design: In addition to migration design centered on AWS multi-cloud, we design architectures that withstand high load and high transaction volume, scaling up without strain to a production-grade setup.
• Observability and quality guardrails: We embed mechanisms to measure and compare behavior before and after migration, along with monitoring and observability, to support safe and explainable modernization.

Why Amatsukaze Is the Right Partner ​

The success of legacy modernization hinges not simply on whether AI can convert code, but on whether you can integrate code comprehension, migration execution, and post-migration operations into a single coherent strategy. Amatsukaze's strength is a structure that lets us work alongside you across all three.

• Infrastructure operations expertise: We bring knowledge cultivated through the design of high-load, high-transaction systems and the construction and operation of robust infrastructure. We support the hardest phase of all — "rebuilding while keeping it running" — grounded in real operational experience, not theory on paper.
• AI / LLM engineering: We have hands-on capability with the latest AI coding agents, LLM-driven code analysis, and RAG, applying hands-on expertise in legacy-asset analysis and migration to production-grade solutions.
• End-to-end partnership: We support everything from current-state analysis and strategy through phased migration and post-migration cloud operations and observability under one roof. Rather than "rebuild and done," we work with you to build a mechanism that keeps generating value.