Hundreds of subsidiaries worldwide of a leading Japanese conglomerate are using FPT’s AI Agents as “digital colleagues” every day across specialized business processes, helping reduce work processing time by up to 90%.
The Challenge of Cross-Border Operations
With hundreds of subsidiaries operating worldwide, the Japanese conglomerate faced growing pressure to maintain seamless operations across time zones while optimizing costs. The challenge was not simply to work faster, but to address three major operational bottlenecks: a growing volume of manual tasks, fragmented data across multiple systems, and language barriers in a multinational environment.
According to FPT’s implementation team, many AI projects typically begin with a general-purpose chatbot designed to address these three operational challenges. However, this approach was not sufficient to meet the customer’s needs, as each department has different workflows, data, terminology, and control standards.
FPT therefore adopted a different approach: breaking down the challenges by business function and developing AI Agents based on a multi-agent model. Each agent is specialized for specific tasks, such as data retrieval, document processing, or translation, while multiple agents can collaborate on a unified platform.
This approach enables businesses to automate repetitive tasks, connect and leverage fragmented data, reduce language barriers, and standardize processing methods and output controls across different units.
“We designed the system so that even employees who are least comfortable with technology can use it with ease. There is no need to learn coding or follow complicated processes. Users simply ask a question, and the ‘digital colleagues’ automatically collaborate, retrieve information, and analyze data to provide an answer,” said Mr. Nguyen Thanh Phong, Director of the Enterprise Services Business Development Center in Japan, FPT Corporation.
FPT introduces technology solutions to more than 300 Japanese customers at an event in Ha Long.In finance and commerce, for example, AI Agents support invoice verification and document reconciliation, identifying multiple types of formatting and content errors and significantly reducing review time.
In logistics, AI Agents help optimize delivery planning and balance inventory levels with transportation capacity based on operational data, enabling teams to make faster decisions in response to changing conditions.
In communications and marketing, AI consolidates market signals from multiple sources, identifies emerging trends, and recommends content, shortening the cycle from idea generation to execution.
For day-to-day operations, AI Agents can record conversations, convert speech to text, translate content, and automatically generate meeting minutes, making the technology particularly useful in multilingual environments and for teams working across different time zones.
Breaking the Limits of Traditional Operations
Following implementation, the system delivered measurable results. Work processing time decreased by up to 90%, with tasks that previously took hours completed in just minutes. Error rates fell by 80%, supported by automated verification, cross-checking, and standardized output. Translation accuracy reached approximately 95%, even for complex language pairs such as English and Japanese.
FPT engineers deploy a global-scale AI system for the customer.The system is now deployed across hundreds of subsidiaries worldwide, supporting a wide range of business functions.
Beyond efficiency gains, however, the greater value lies in transforming how the organization makes decisions and operates.
Previously, requests to modify product features often had to be finalized several months before launch because of the time required for processing and internal coordination. With “digital colleagues” supporting specialized business processes, the conglomerate can now receive and process change requests much closer to the launch date, reducing the lead time from months to just days.
The ability to automate and coordinate these business processes also reflects Level 3 of CASAN, FPT’s framework for measuring enterprise AI-Native maturity, where AI begins to participate directly in operations rather than merely supporting individual tasks.
AI is not only helping the enterprise work faster but also creating a more agile operating foundation. The technology enables faster decision-making, reduces the time and coordination steps required between units across multiple countries, and allows employees to focus on higher-value work.