Fujitsu’s automated code AI cuts legacy chaos

Fujitsu’s latest move brings a simple but powerful idea to life—what if legacy code could explain itself? With its new Automated code documentation AI, the company is tackling one of the biggest challenges in enterprise IT: understanding old systems that few people can still read.

The new service, Fujitsu Application Transform powered by Kozuchi, analyzes source code and generates design documents automatically. This changes the starting point for modernization. Instead of spending weeks trying to understand systems, teams can now begin with clear, structured documentation.

A faster path to modernization

Legacy systems, especially those built on COBOL, often slow down innovation. They are complex, poorly documented, and heavily dependent on expert knowledge. This is where Legacy system modernization with GenAI starts to show real value.

By reducing documentation work time by nearly 97%, the solution removes one of the biggest bottlenecks in modernization projects. It also improves the quality of output, delivering more complete and readable design documents compared to traditional methods. For organizations, this means faster decision-making and smoother transitions.

How the AI actually works

What makes this approach stand out is not just automation, but accuracy. Fujitsu combines its own code analysis technology with generative AI, supported by a knowledge graph-based RAG system. This ensures that the AI understands relationships across large volumes of source code.

The result is documentation that avoids common AI issues like missing details or inconsistencies. Even complex systems are translated into structured, easy-to-follow designs. This helps teams move forward with confidence, without constantly second-guessing the output.

Reducing dependency on experts

Traditionally, understanding legacy systems required highly experienced engineers who knew specific programming languages. That knowledge gap often delayed projects and increased costs. With this new approach, even teams without deep expertise can interpret system structures.

The ability to reduce system documentation time with AI also changes team dynamics. Instead of relying on a few specialists, organizations can distribute work more evenly and speed up execution. It’s a small shift in process, but a big shift in capability.

What comes next

Fujitsu is not stopping at documentation. The roadmap includes features for rewriting source code, rebuilding systems, and supporting ongoing operations. This signals a broader move toward end-to-end modernization powered by AI.

The early response, including feedback from enterprise users, points to strong potential. As organizations continue to deal with aging systems, tools like this could become essential rather than optional.

The takeaway

The Automated code documentation AI approach is not just about saving time. It changes how organizations think about legacy systems—from something difficult to manage to something easier to understand and evolve.

For enterprises looking to modernize, the message is clear. When understanding improves, everything else moves faster

Read More: 

Crowdstrike and HCL strategic partnership reshapes cyber defense

GIGABYTE Z890 Plus Series lands in India with one-click 40% performance boost

boAt Malaysia launch 2026 taps digital consumers

MongoDB APAC Strategic Partner Program reshapes AI adoption playbook in Asia Pacific