How Fujitsu Rebuilt 4,000 Legacy Systems Around AI Operations

When Fujitsu began confronting its legacy infrastructure problem in 2020, the company was managing over 4,000 fragmented systems, with 300 employees needed just to process 20,000 contracts a year. Today, AI agents handle contract verification, compliance checks, and reconciliation — and 80,000+ employees are using AI-enabled tools daily. The transformation is a case study in what enterprise AI integration actually looks like at scale.

The Problem: Japan's "2025 Digital Cliff"

Fujitsu's challenge mirrors that of Japanese enterprises broadly. Policymakers coined the term "2025 Digital Cliff" to describe the risk that ageing systems, rising maintenance costs, and shrinking technical workforces could cost Japan up to US$65 billion annually in lost productivity after 2025. Legacy ERP environments — customized, poorly documented, and understood only by engineers who've long since retired — are at the center of this problem.

Global IT competitors including IBM, Accenture, and cloud-native software providers had already accelerated toward AI-enabled operations, putting traditional IT services firms like Fujitsu under direct pressure to evolve beyond maintenance contracts.

The YLP Plus Transformation Programme

Fujitsu's CDTO Kohei Toyama launched YLP Plus to create a "global single instance" — a unified architecture replacing fragmented regional systems with common standards and real-time visibility.

Key moves in the overhaul:

• SAP S/4HANA migration standardizing applications, data, and workflows globally
• SAP Build Work Zone consolidating access to 750 enterprise applications for 16,000 employees
• AI agents and generative AI tools taking over contract verification, compliance, and management reporting
• Executive meetings now powered by AI-generated simulations, dashboards, and benchmarking data

The "Customer Zero" Strategy

Perhaps the most actionable insight from Fujitsu's experience is its "Customer Zero" approach: deploying and stress-testing technologies internally before selling them to clients. This directly addresses the credibility gap that many AI vendors face — enterprises are increasingly skeptical of transformation claims that haven't been proven in production.

For any organization considering large-scale AI adoption, Fujitsu's experience reinforces one key lesson: modernization is no longer primarily about replacing software. It's about redesigning workflows, consolidating fragmented teams, and redefining how decisions get made.

🔗 Read the full article on iTnews Asia