Fujitsu Leverages LLMs for Automated Legacy System Modernization, Cutting Migration Time by 40%

Fujitsu Limited today announced the Japan launch of its AI-driven Modernization Service. This new offering combines Fujitsu's extensive system integration expertise with generative AI, leveraging both its proprietary AI platform Fujitsu Kozuchi and the Takane large language model (LLM), alongside cutting-edge third-party LLMs such as Anthropic's Claude and OpenAI's GPT. The service is designed to automate and optimize modernization initiatives, specifically focusing on rewrite and rehost approaches for legacy systems. Fujitsu claims this service can shorten migration periods by approximately 40%. For cloud and DevOps practitioners, this announcement is highly significant. Legacy system modernization is a persistent and often daunting challenge, consuming substantial resources and time. Fujitsu's service directly addresses this bottleneck by automating large portions of the process. This means development teams can potentially reallocate significant effort from manual code conversion and refactoring to higher-value tasks like innovation, new feature development, and strategic architectural improvements. The promised 40% reduction in migration time translates directly into faster time-to-market for modernized applications and a quicker return on investment for digital transformation initiatives. It also mitigates the risk associated with complex, large-scale migrations by introducing AI-driven consistency and efficiency. This development fits squarely within the broader trend of applying generative AI to enhance developer productivity and automate complex IT operations. Over the past year, we've seen a rapid expansion of AI-powered coding assistants, automated testing tools, and AI-driven infrastructure management. Fujitsu's offering builds on this by extending AI's reach into the historically labor-intensive domain of legacy system modernization. The integration of multiple LLMs (Fujitsu's own Takane, Claude, and GPT) reflects the industry's move towards multi-model strategies, where organizations select and combine the best-of-breed AI capabilities for specific tasks, rather than relying on a single vendor. This approach acknowledges the varying strengths of different models and the need for specialized AI agents to handle diverse aspects of software development and transformation. Practitioners should view this as a powerful new tool in their modernization arsenal. While the 40% reduction is compelling, it's crucial to understand the service's applicability. It's particularly geared towards sectors with frequent legal revisions like finance, public services, and healthcare, and industries with complex operational know-how embedded in legacy systems, such as manufacturing and distribution. Teams should evaluate how well their specific legacy codebases and target architectures align with the service's capabilities. Key considerations will include the fidelity of the AI-generated code, the ease of integration with existing CI/CD pipelines, and the level of human oversight required for verification and refinement. This service could free up specialized "Modernization Meisters" (as Fujitsu calls them) to focus on architectural design and strategic problem-solving rather than rote conversion, ultimately accelerating the shift to cloud-native and modern architectures.