- Fujitsu promotes technologies for multiple agents of IA collaborate safely between companies and sectors, with a special focus on supply chains and healthcare.
- Services such as Fujitsu Kozuchi AI Agent and AI Auto Presentation demonstrate how AI agents can act as team members, automate presentations, and respond in real time.
- The strategic alliance with NVIDIA It allows Fujitsu to build a complete AI infrastructure, with FUJITSU-MONAKA CPUs, NVIDIA GPUs, and NIM microservices for self-evolving agents.
Fujitsu's bet on the artificial intelligence agents It is redefining how companies interact with data, automate processes, and coordinate with each other in complex environments such as supply chains, healthcare, and manufacturing. Far from simply offering isolated AI models, the Japanese company is building a complete ecosystem where multiple intelligent agents can collaborate with each other and with people in a secure and governed manner.
In recent years, Fujitsu has unveiled several key technologies: platforms for AI agents from different companies to work together, services like Fujitsu Kozuchi AI Agent that act as another member of the team, avatars capable of creating and presenting presentations Available in over 30 languages, and backed by a powerful alliance with NVIDIA to support it all with a high-performance computing infrastructure. This entire initiative has a clear objective: to accelerate the adoption of AI in the business and social fabric while always maintaining the autonomy and control of organizations.
Technology for multiple AI agents to collaborate across companies
One of Fujitsu's most significant innovations is the development of multi-agent AI collaboration technologies originating from different companies and suppliers. This capability is especially critical in environments such as the supply chain, where manufacturers, distributors, pharmaceutical companies, logistics operators, and many other actors need to coordinate without compromising the security or confidentiality of their data.
Fujitsu's proposal allows that AI agents from different organizations cooperate securelyexchanging information and reacting quickly to sudden changes in the environment. We're talking about situations like an unexpected spike in demand, a supply disruption, a natural disaster, or a critical production incident at a key supplier.
To demonstrate the potential of this technology, Fujitsu will begin field trials with Rohto Pharmaceutical and the Tokyo Institute of Science (Science Tokyo) from January 2026. The goal is to optimize Rohto's supply chain, improving responsiveness, reducing inefficiencies and ensuring agile recovery from unforeseen events, something especially sensitive in the pharmaceutical sector.
This initiative is also part of the activities of the Council for Competitiveness-Nippon (COCN)Fujitsu will participate in this initiative to promote AI spaces that enable companies to collaborate by securely sharing data and intelligence. The aim is to strengthen the competitiveness of Japanese industry through AI agents operating in multi-company ecosystems while maintaining a focus on governance and trust.
In the medium term, Fujitsu plans extend this technology to broader and more complex supply chainsand offer it within its Dynamic Supply Chain services, integrated into the Uvance business model. The goal is to help organizations design much more resilient supply chain strategies, capable of operating sustainably despite global volatility, and based on cooperation between AI agents that cross organizational and sectoral boundaries.
Fujitsu Kozuchi AI Agent: AI as another member of the team
On the corporate front, Fujitsu has launched Fujitsu Kozuchi AI AgentThis service is designed to enable AI agents to perform high-level tasks both autonomously and in coordination with humans. It's not just a simple assistant; it's an agent capable of understanding complex objectives, proposing approaches, and executing plans using various AI models.
Thanks to a Fujitsu patented processing logicKozuchi AI Agent is capable of breaking down abstract questions raised in a conversation into concrete and manageable tasks. From there, it generates an action plan, selects the most suitable AIs for each task, gives them specific instructions, and formulates a proposal or solution based on the information gathered.
The service integrates both Fujitsu's own AI models such as Takane and Kozuchi AutoML Like other external models, Takane stands out for its very high level of Japanese language proficiency and extensive customization options, while Kozuchi AutoML automates the generation of advanced machine learning models in a very short time, making it easier for companies to create tailored solutions without huge teams of data scientists.
All of this is offered through Fujitsu Data Intelligence PaaSFujitsu Uvance, the company's all-in-one data operations platform, is integrated into the Fujitsu Uvance business model. This platform acts as the backbone for deploying AI agents, connecting disparate data, and orchestrating complex workflows across various business areas.
The first sales agent in this family has focused on debates on business profitability and trade negotiationsThis agent is designed to add relevant information, present key indicators, and suggest measures to improve the company's position during negotiations. Building on this foundation, Fujitsu plans to deploy new agents specializing in production management, legal matters, and other corporate areas where intelligent automation can provide a competitive advantage.
Fujitsu has also announced a wider deployment of Kozuchi AI Agent during fiscal year 2024Fujitsu Uvance is integrating this technology into its digital offerings, including the Work Life Shift program. In this way, the company aims to make collaboration between humans and AI agents a part of the daily operations of organizations, from decision-making to time and task management.
AI agent platform for Japanese healthcare
The healthcare sector is another major area where Fujitsu's strategy is being applied. The company has developed a specific AI agent platform for Japanese healthcare, with the aim of improving operational efficiency, ensuring the continuity of medical services and relieving the workload of clinical and administrative staff.
The heart of this proposal is a orchestrator or coordinator AI agentThis system acts as a centralized platform to coordinate multiple agents specializing in different tasks within the medical workflow. These agents can be developed by Fujitsu or its technology partners, and encompass functions such as structuring clinical data, monitoring interoperability, managing appointments, supporting coding and billing, and assisting with patient triage, among others.
The platform is designed so that AI agents collaborate seamlessly with each otherCombining its capabilities to transform end-to-end healthcare operational processes, both within medical centers and in interactions with other actors in the system (laboratories, insurers, administration, etc.). The coordinating agent is able to control and automate complex workflows, connecting various specialized medical applications without requiring the professional to switch between systems.
With this approach, Fujitsu intends that healthcare professionals can focus on high-value tasks such as diagnosis, clinical decision-making, and direct patient care. Center managers will be able to reassign staff to critical roles, improve overall productivity, and offer better working conditions—key to retaining talent and reducing burnout.
Patients, for their part, will benefit from shorter waiting times and more personalized servicestailored to their needs and circumstances. By having more coordinated and automated workflows, the healthcare system can react better to peak demand, manage appointments more accurately, and reduce errors resulting from fragmented manual processes.
This platform for AI-powered healthcare agents is being developed with NVIDIA's technological supportFujitsu brings its expertise in accelerated computing and key technologies for AI agents, such as NVIDIA NIM microservices and NVIDIA Blueprints. The idea is to combine Fujitsu's deep knowledge of medical information systems in Japan with the power of NVIDIA's AI infrastructure to enable cutting-edge medical operational practices globally.
Looking ahead to 2025, Fujitsu plans accelerate the commercialization of this platformCollaborating with leading medical institutions and international partners to validate the efficacy of the coordinating agent and develop new, specialized agents for different areas of healthcare. Under the Uvance umbrella, the company aims to continue transforming both healthcare and drug discovery through the intensive use of data and AI agents, with a focus on increasingly personalized medicine and improving people's well-being.
Fujitsu AI Auto Presentation: Avatars that present and respond in more than 30 languages
Another of Fujitsu's most striking developments is Fujitsu AI Auto PresentationThis technology allows AI-powered avatars to create, deliver, and adapt presentations in a fully automated way. This solution is part of the Fujitsu Kozuchi AI service and integrates as an agent within the ecosystem of Microsoft 365 Copilot.
The tool is capable of Generate and deliver presentations from Microsoft PowerPoint filesby extracting content from the slides and constructing a coherent speech in the user's chosen language. But it doesn't just read text: the avatar can answer audience questions in real time, drawing on pre-integrated materials thanks to a Retrieval Augmented Generation (RAG) approach, allowing it to provide answers based on corporate or technical documentation.
A differentiating factor is the possibility of create personalized avatars with the user's own image and voiceUsing speech recognition technologies, large language models (LLM), and speech synthesis, the avatar not only delivers the content but does so with a natural tone that closely resembles the style of the person it represents. Furthermore, the tool supports over 30 languages, facilitating communication in international settings without requiring the speaker to be fluent in every language involved.
Fujitsu AI Auto Presentation incorporates a Patented autonomous slide transition function with timing controlThe system generates the text to be displayed based on the number of characters on each slide and the time restrictions set for the presentation, automatically changing slides at the appropriate time to fit the planned duration without losing fluidity.
The solution also allows a granular customization of the presentation contentSlide by slide, prompts can be added to fix certain text, generate additional content, insert specific information, or adjust the writing style (more formal, more direct, sales-oriented, etc.). This flexibility makes it possible to reuse the same slide deck, quickly adapting it to different audiences or contexts.
This development has been carried out jointly with Headwaters Co., Ltd.Fujitsu, a company specializing in AI solutions, will offer its "Fujitsu AI Auto Presentation" AI agent within the Microsoft 365 Copilot declarative agent framework. Fujitsu will begin using this technology internally starting in the second quarter of fiscal year 2025 and plans to roll it out to customers worldwide starting in the third quarter, with plans to integrate it directly into Microsoft 365 Copilot. Microsoft Teams and PowerPoint.
The purpose of this tool is democratize the process of creating and presenting presentationsPeople with limited time, no public speaking experience, or language barriers can delegate the delivery of their message to an avatar, ensuring a professional and consistent level of presentation. According to Fujitsu, this will help organizations share high-quality information, gain operational efficiency, and move toward a more inclusive digital society, in line with their materiality strategy.
The official announcement has been supported by executives from Microsoft Japan and HeadwatersMicrosoft highlights that the integration of Fujitsu AI Auto Presentation into Microsoft 365 Copilot will help companies streamline the creation of multilingual content, optimize marketing operations, and facilitate faster knowledge sharing both within and outside the organization. Headwaters emphasizes that the project demonstrates the practicality of AI agents integrated into real-world business processes and their enormous potential to improve operational efficiency, with a commitment to further democratizing these agents.
Fujitsu-NVIDIA strategic alliance for a complete AI agent infrastructure
To support this entire deployment of intelligent agents, Fujitsu has strengthened its strategic collaboration with NVIDIA With the goal of building a high-performance AI infrastructure that natively integrates AI agents across multiple sectors, this alliance seeks to offer a robust technology platform that allows companies to leverage AI while maintaining their autonomy in managing data, models, and operations.
The collaboration focuses on the joint development of an AI agent platform and a state-of-the-art computing infrastructureOn the one hand, work is underway on a platform adapted to specific sectoral needs (healthcare, manufacturing, robotics, among others), supporting multi-tenancy environments where different organizations can use agents securely and in isolation, but sharing an efficient technological base.
To achieve this, Fujitsu is integrating its AI workload orchestration technology, based on its AI computing brokerage, with the NVIDIA Dynamo platformBased on this, the aim is to create mechanisms that allow AI agents and models to evolve autonomously and be customized according to the sector and the client, leveraging NVIDIA NeMo and improving Fujitsu's multi-AI agent technologies, including the optimization of the Takane model.
The agents developed in this environment will be supplied as NVIDIA NIM microservicesdesigned to deliver highly optimized inferences, this will make it easier for businesses to adopt advanced AI agents without having to deal with the complexity of deploying and scaling large models on their own, accelerating the AI industrial revolution with agents that continuously learn and improve.
In terms of infrastructure, the alliance includes the creation of an optimized AI computing platform from the silicon levelThis system integrates the FUJITSU-MONAKA CPU series with NVIDIA GPUs via NVIDIA NVLink-Fusion technology. The goal is to achieve near-zero-scale performance, enabling widespread industrial adoption in compute-intensive fields such as scientific simulation, advanced manufacturing, and autonomous robotics.
This infrastructure will include a complete HPC-AI ecosystem, combining Fujitsu's high-speed software for ARM with NVIDIA CUDAThanks to this, the aim is to offer comprehensive support for the entire AI lifecycle: from training and tuning models to deploying and operating agents in production, including monitoring and continuous updates.
The agreement also places a lot of emphasis on the creation of a robust ecosystem of partners and customersFujitsu and NVIDIA aim to launch joint programs with companies, institutions, and developers to expand the use of AI agents and models, generating transformative use cases in sectors such as manufacturing, healthcare, energy, and robotics. Among the initial applications are digital twins to accelerate manufacturing processes, robotics to address labor shortages, and advanced automation systems based on physical AI.
The management of both companies has emphasized that The AI industrial revolution is already underway And that it is urgent to build the necessary infrastructure to support it. Fujitsu, with its track record in supercomputing, quantum research, and enterprise systems, is positioned as a key partner for NVIDIA in Japan and beyond. Together, the two companies aim to establish this AI infrastructure as a cornerstone of Japan's digital society by 2030, with the goal of later expanding it to other global markets.
In strategic terms, Fujitsu frames this alliance within its objective of to promote a sustainable social and business transformation, aligned with the Sustainable Development Goals (SDGs). By offering a cross-cutting infrastructure for AI agents, the company aims to help drive innovation in Artificial Intelligence It should not be limited to large corporations, but should extend to sectors and organizations that until now had excessively high barriers to entry.
With over 113.000 employees worldwide and consolidated revenue of several trillion yen, Fujitsu remains the Japanese leader in digital servicesTheir strategy combines five key technological pillars—AI, computing, networking, data and security, and convergent technologies—with a clear purpose: to build social trust through innovation and underpin sustainable digital development. The collaboration with NVIDIA, along with the AI agent platforms described, fits perfectly into that purpose.
Taking all these initiatives together—collaboration between agents from different companies, services like Kozuchi AI Agent, healthcare platforms, avatars for presentations, and a complete AI infrastructure with NVIDIA—, Fujitsu is building an ecosystem in which AI agents become central to economic and social activityThe vision is not limited to automating isolated tasks, but to orchestrating complex collaborations between humans and machines, between organizations and sectors, relying on a powerful, secure and governed technological base that aims to set the pace for the new era of applied artificial intelligence.
Passionate writer about the world of bytes and technology in general. I love sharing my knowledge through writing, and that's what I'll do on this blog, show you all the most interesting things about gadgets, software, hardware, tech trends, and more. My goal is to help you navigate the digital world in a simple and entertaining way.