KEY POINTSHitachi launches Agentic AI Integration Platform for enterprise system integration from planning and requirements to testing and operationsInternal results show about 200-fold productivity gains in design-to-test work, with a 30% companywide process target for fiscal 2027Platform combines frontier AI models, Hitachi system-integration knowledge and customer-specific enterprise context for mission-critical projects
Hitachi’s new AI-driven system integration platform. Note: Japanese text in the original image from press release has been translated into English by our editorial team.
Hitachi Ltd.
Hitachi Ltd. has developed an AI-driven platform for enterprise system integration and plans to apply it across growing modernization demand, targeting a 30% productivity improvement across the full development process by fiscal 2027, according to a July 24 release.
The platform, called Agentic AI Integration Platform, applies AI across the system integration process from external environment analysis and initial planning to requirements definition, design, coding, testing and operations. Hitachi said the platform embeds task-specific AI agents for each stage to improve productivity while maintaining safety standards for large-scale and mission-critical systems.
Internal results showed roughly 200-fold productivity gains from the design stage through testing, while productivity in requirements-definition work improved by as much as 240 times. Hitachi plans to promote adoption in large-scale mission-critical projects, including sectors with sizable legacy systems such as finance, public services, energy and railways.
A central feature of the platform is a knowledge base that stores customer-specific management intent, operational know-how and decision criteria as what Hitachi calls “enterprise context,” alongside the company’s own system integration knowledge. AI agents automatically accumulate and update that knowledge as projects move through planning, development and operations, allowing the system to improve autonomously while shortening development periods, the company said.
The platform combines frontier AI models under global partnerships with Hitachi Group’s IT and operational technology system integration expertise and AI technology from GlobalLogic. It allows flexible use of models from Anthropic, Google Cloud and OpenAI, and includes Hitachi’s own generative AI development framework and GlobalLogic’s VelocityAI. In one linked environment, VelocityAI reduced token usage costs by as much as 54% compared with standalone AI use. Hitachi has also begun internal trials of an OT-specific code-generation base compatible with Anthropic’s Claude and the latest OpenAI models.
Hitachi’s Forward Deployed Engineer, or FDE, teams will work at customer sites during planning, development and operations as AI transformation partners, the company said. The approach is intended to help clients speed business transformation while preserving the safety and quality requirements of core enterprise systems. In Japan, mission-critical systems are widely used in regulated industries and public infrastructure, where outages or defects can have broad operational consequences.
The company disclosed examples of external validation in specific projects. In core insurance system development with MS&AD Systems Co., coding and unit-testing productivity improved by about 25%. In work with Shizuoka Bank and Shizugin IT Solutions on an open mission-critical core banking system, coding and unit-testing productivity improved by about 30%, and practical deployment is under way.
Hitachi said AI is already used in more than 60% of its system development work. The company established the Hitachi Application Framework Justware in 2002 and has applied high-quality and safety-focused development practices in more than 200 large projects in finance, public and social infrastructure fields.
Hitachi posted revenue of 10.5867 trillion yen ($64.6 billion) in fiscal 2025, ended March 2026, and had 606 consolidated subsidiaries and about 290,000 employees worldwide as of March 31, 2026. The company has expanded AI partnerships in recent months, including a strategic partnership with Anthropic on May 19, a broader alliance with Google Cloud on June 9 and deeper collaboration with OpenAI on June 17. Exchange rate used: USD/JPY = 164.