We are now beginning Fujitsu Uvance update 2025 to introduce the business progress of Fujitsu Uvance to the media, investors, and analysts. Now, let's invite our first presenter, Yoshinami Takahashi, our Corporate Executive Officer, Corporate Vice President, and COO. Takahashi-san, over to you. Hello, this is Takeshi speaking, and thank you very much for attending Uvance update today. Last year, in Uvance update, I mentioned that in Uvance, we will utilize our strengths of data and AI to imagine creating both business impact and social impact. I shared with you this commitment. One year has passed since then, and through co-creation with many customers and partners under the Uvance brand, we've been working on various use cases and solutions to promote DX and SX as our workplaces. Last fiscal year, Uvance revenue target was JPY 450 billion, but we were actually able to achieve JPY 482.8 billion, exceeding our target. This fiscal year, we are aiming at JPY 700 billion. Through Uvance service solution, Uvance has grown to account for 30% of Fujitsu service solution revenue, and it has become a business that drives our own business transformation within Fujitsu. Today, we would like to introduce the evolution of AI, through examples of data and AI, as well as the use of agents in various business use cases, to share with you how we want to impact society. Now, during the past one year, there's been significant evolution in AI. Until last fiscal year, AI was still in the trial phase, as we understand, but we have now transitioned to the phase to generate outcomes. In fact, in a survey of 800 CXOs in 15 countries conducted by our company, the adoption rate of generative AI, GenAI, reached 98%, and as of which, 60% of companies that have adopted it said that their productivity has increased by 10% or more. We believe that GenAI will be promoted more, and through the collaboration of people in AI, we think we can promote this further. According to various interviews and so forth, by 2030, more than 80% of companies predict that AI will be incorporated into more than 50% of their processes. AI is now taken for granted. Management decision, automation of business, for all of this, AI can support, and AI can therefore support various corporate activity. AI is now collaborating with people. With regards to Uvance as well, there's evolution. During the past one year, we've been able to go through a lot. We conducted this demo last year, and a global inventory can be visualized in just a few steps, and we can visualize the inventory. We can minimize CO2 emission, and the supply chain can be optimized at the time of emergency. That was the kind of demonstration we showed to you last year. We are going to evolve this further. We will embed AI agents, and from people prompting, AI agents will autonomously operate. And the judgment process has been done by humans, but this will be done by AI agents, which will increase the speed and accuracy of decision. When we say AI agent, and we're speaking in meetings, doing research, and so forth, AI agents are now often used. What we are aiming at with regards to AI agents is to go one step further, and for specific business issues, we want to develop a specified AI agent to solve various issues in specific workplaces. And I think it's easiest to understand to introduce you to a specific use case. We will now introduce you to specialized agents that are being developed by Fujitsu. Mr. Hiroki Hiramatsu, please Hello. This is Hiroki Hiramatsu speaking from Fujitsu. I want to introduce to you the AI agent I've been responsible for developing it. The theme for this year was the same as last year, "Stockout Risk." AI agents will think and make decisions instead of human beings, and human judgment can be supported by AI agents. That is how we've been able to evolve. I will now stop my demonstration, and the time is limited, but I hope you can get an understanding of how AI agents work. A certain product is likely to suffer from a stockout. That is where the story starts in the demonstration. The user's policy is to place emphasis on cost. The user policy is understood by AI agent, and the AI agent autonomously starts to operate. And in order to realize this mechanism, four types of specialized AI agents and two types of comprehensive AI agents are used. Behind this screen, there's this inventory agent, production agent, and sales agent. They all incorporate the necessary data, and real-time calculation is made, and from various options, the best option is chosen, and these cards from left to right, they move, and once all of the proposals or options are available, as you can see in the middle of the screen, these specialized AI agents are orchestrated by a team leader type of AI agent, which chooses the best option, and the evaluator, as the word denotes, judges whether it can automatically be approved or whether human judgment is necessary, and furthermore, as a result of this mechanism, conventionally, many people, as well as time, had to be allocated to discuss and consider. But in just a couple of seconds, in dozens of seconds, proposals can be reached using this AI agent mechanism. Without doing anything, AI agents autonomously operate, and as you can see on the screen, the optimum, the best option is selected. But in accordance with necessity, action required button comes up so that human beings can intervene to make further decisions. So AI moves first, and the brake and the handle, that's under the responsibility of the people, and I think that will be the best solution for various workplaces. Now, behind this screen, the sales agent is making various considerations. And time is limited today, so the stockout alert, how it's recognized and how judgment is made, that is what I want to show to you today. And for the stockout risk, a stock agent, inventory agent, chose the best option automatically, and it's automatically approved. That's not the end of the process. The user for each agent, what is discussed amongst various agents, and why was this conclusion reached? The dialogue between and among AI agents can be taken a look at over here. The history, the record of discussion, it can be taken a look at, and the scores and the comments can be used to feedback by human beings. Another interesting point is user feedback appears in the top right corner. There's the governance agent. The governance agent takes a look at the feedback real time, and the scoring is done for each AI agent, and procurement agent in purple. User feedback mentions that the proposal was quite ambiguous. The governance agent not only scores or evaluates, but analyzes the weakness, tries to make improvements, so that next time, a better proposal can be made. The decision, as well as the thinking of the AI agent, the prompt part, is automatically updated. In other words, AI is not just implemented, and they continue to be trained and learn. Just like human beings, the more experience you go through, the more improvement you can achieve. And as you just saw, people and AI agents, they are collaborating, and we are now in this time. And I introduce to you today the supply chain use case, and it's a great honor for us to be able to introduce this use case. In the demonstration time later on, I can explain in more detail, and I can show to you real time how the cards move. So please feel free to take a look at the demonstration later on as well. Thank you very much Mr. Hiroki Hiramatsu, thank you very much. The solution that you just took a look at, it was written in English, and the Japanese language is also available. It's a state-of-the-art technology or solution where people and AI can collaborate, and in the World Economic Forum, AI Governance Alliance Minds program, this solution was adopted there, and globally, only 18 companies worldwide were selected. It was highly evaluated internationally as an innovative use case that will have an impact on society. Furthermore, in the field of generative AI, GenAI as well, we've been recognized as a leader by global analyst firms. The first time as a Japanese company, we were recognized for our implementation capabilities and reliability, attracting significant attention. These cutting-edge initiatives are not limited to the supply chain. They're also being implemented in Uvance's vertical areas, which aim to solve vital issues across industries in a cross-industry manner. We would like to introduce some cases to you today. First, last year in Uvance update, we announced a strategic partnership with Cohere. The result of this joint development is Fujitsu's unique LLM, Takane. Takane's features include high accuracy, Japanese language understanding, and a secure private environment. Its introduction is progressing in a confidential, sensitive field, such as medical care and so forth. And we would like to highlight today the Takane use case in the medical industry. Currently, in medical settings, there is a large amount of important information, such as medical records and test reports that are essential for treatment and research. And one issue is that this data and reports, they're written by physicians in free form. As a result, if we try to analyze this data, it takes a massive amount of time and effort. As a result, the clinical data is not utilized to its fullest extent. As a result, Fujitsu decided to collaborate with Tohoku University to tackle the challenge of structuring medical data. Even with Japanese, which is difficult to understand in context, Takane achieved an accuracy rate of over 80% in verification using real data in a secure private environment, surpassing conventional generative AI and target values. 80% means that we've been able to structure medical data, and side effect prediction and so forth are now possible. By utilizing data, we can realize personalized medicine that suits the needs of each patient, and we've been able to make a significant step forward with the realization of personalized medicine. The next case study is in a little bit different industry. Centered around data scientists, we are working on the infusion of AI together with customers. That is what I would like to introduce next. This is a case study from Mazda, the automobile manufacturer. Mazda is promoting a company-wide DX-based reform of its business structure to improve business efficiency. Using our data platform, in just two years, they have deployed 33 business applications in five departments. In factories, sales, purchasing, and IT, our data platform is being used for data-driven decision-making. How did they, or we, achieve such a company-wide transformation? For example, in the purchasing department, we normally hear that data is not centralized. So we first sort out the issues, and we visualize the result. And against that integrated data, we consider how we can utilize such data. Based on such ideas, the customer eventually puts together a DX concept for purchasing on their own. Currently, we discuss with customers in the field and closely support them so that they can take the lead in their own transformation. We have data scientists internally. I believe that such customer-facing data scientists are dispatched to companies to support them. And that is a critical service of Uvance. And when data and AI further penetrates, we need to accelerate the fusion of data and AI going forward. Another case study is, well, the society or social environment around us is rapidly changing. Unexpected events such as geopolitical risks, natural disasters, and recent tariffs have become a significant impact on businesses. How quickly we can come up with countermeasures is essential, especially in supply chains. We have been seeing disruption of global division of responsibilities. This may lead to a rise of procurement costs as well as the end price. This can directly affect the revenue. The ability to respond quickly to these changes and provide effective solutions is a major value proposition of Uvance. It is the concept of Uvance. The case study I would like to introduce next is from ADEKA, a global comprehensive chemical manufacturer. In ADEKA, they have established a mechanism to immediately identify the impact of external risks on earnings, such as contingencies or tariffs, enabling flexible and responsive decision-making through simulation. Furthermore, in Fujitsu, based on such solution, we are using agentic AI so that we can make the system autonomous. Today, we're joined by Mr. Shirozume, President of ADEKA, who will introduce their initiative through a dialogue with Fujitsu's project manager, Doi San. Now, Shirozume, Doi San, please come to the stage. Mr. Shirozume, Mr. Doi, please come up on the stage. Once again, I would like to introduce Mr. Hidetaka Shirozume, Representative Director, President, and CEO of ADEKA. My name is Shirozume. Thank you for having me today. Shirozume San, thank you for taking time out of your busy schedule today to join Uvance update. In just two weeks, we covered the supply chain data starting from the upstream to the downstream to identify the tariff impact on P&L and also to come up with actions of countermeasures. I think that was very speedy. But what was the reaction or feedback internally, and what was your impression? Yes, it was amazing. Trump tariffs, we have been hearing about it. So for a couple of months, we spent time to prepare for the upcoming changes. But I was surprised how quickly actions were developed through this solution. Thank you. This is true in Fujitsu Uvance and also our Wayfinders Consulting and Data Engineering capabilities. They are also utilized for data-driven management. What do you think about Fujitsu's ability and Uvance, and what are your future plans in utilizing them? Do you have any proposals? Yes. Doi San, in your family, have you ever hit your toenail with the foot of the furniture? That is exactly what we experienced this time. I mean by we are not accurately understanding where our body part is. We often hit our toenails because we don't know where they are. When we came to utilize Uvance, when we looked at our product supply chain, we needed to make an accurate input in order to precisely identify the impact of the Trump tariffs. We have thousands of products that we develop and produce, which uses tens of thousands of raw materials, which are used in plants around the world. They are in multiple places, and we are taking a phased production approach. Such context needs to be input accurately. Just like human body, we still had some parts that we don't know what they are and where they are. That is what we learned by utilizing Uvance. That also means if we know where the small parts of our body are, we will be able to maximize the capability of Uvance. That is the potential we felt through this journey. Thank you. The value of Uvance has been maximized, and customers who are benefiting from such strength are the customers who are having the accurate understanding of who they are and what they have. Yes, in calculating payrolls or bonuses, I believe we'll be able to utilize the solution, so I'd like to extend our deepest gratitude. Recently, we finished the final report on the project, and I heard more about it today, and I have even higher expectations for its future development. I look forward to continuing to work with you. Thank you very much for your time today. Mr. Shirozume, Doi San, thank you very much. Please come off the stage. Mr. Shirozume, Doi San, thank you very much. Well, I also hit my big toe recently. That really hurt. Why does Uvance care so much about achieving significant change in a short period of time? When we look at the social environment, there are many things that are changing very rapidly, so we need to do prototyping and achieve results promptly. By repeating prototyping, we can create impact with speed. In order to guarantee such speed, there are three Rs that we place importance on: real user, real problem, and real data. First, there are real users who are facing challenges. Real problem is the essential issue to be solved in the society. Normally, we use dummy data for analysis. But it isn't worth it unless we use real data. What we tell the data scientists in the front is that they should practice these three Rs. We need to go to real users to solve real problems with real data. In the world of data and AI, these three Rs are going to be critical. Such attitude is practiced by our data scientists. We support our customers' transformations, and that is what Uvance exists for. Now, let me introduce you to our further efforts in utilizing AI agents. We are a system integrator, so in the area of ASI and system development and maintenance, we are thinking of how we can utilize AI agents. In various domains, internally, we are using business-specific agents. Now, we will introduce the project we're working on with SBI Sumishin Net Bank. First, I would like to explain the background and intent of this project, and I would like to call in Aikawa-san, who's responsible for the project. Yoshinami-san, thank you for the introduction. Hello, everyone. As introduced, as the evolution of AI accelerates, Fujitsu's strong area of systems development has reached a major turning point. For example, GenAI is used for writing code or automatically generating macros, creating new things. The use of AI is expanding in such areas or source code generation tools. We already have such tools. The entire system development, it's not just used for programming or code writing. That is not enough. What we think is we shouldn't be stuck in partial optimization, but we need to not only improve but evolve system development practices. That's what the management needs to accomplish. On the other hand, many companies already have a legacy system that's been in place for five or 10 years, which is complicated, and they are connected to other systems. It is their worry how they can continue using those systems and making transformation at the same time. One of the key points is that they are not built by themselves. They may be dependent on certain people. There may be lack of documentation, and huge man hours and costs are required for modification. With the evolution of AI, we want to fundamentally change these normal expectations. We worked with SBI Sumishin Net Bank to demonstrate the use of AI agents in the maintenance of their existing systems for the sake of transformation. I will ask Mr. Aikawa about the system that plays a central role in their business and the potential of AI. We will get the stage prepared. If you could please wait a while. Thank you very much. Mr. Aikawa, could you please come on stage? Aikawa-san, thank you very much indeed for coming today. We would now like to start the talk session. Joining us today is Mr. Shinichi Aikawa, Executive Officer, Head of System Division, SBI Sumishin Net Bank. This is Aikawa speaking. Thank you very much for this precious opportunity, and Mr. Aikawa has a vision of fundamentally transforming system development through generative AI. He has this grand vision. Towards this vision, SBI Sumishin Net Bank is working with Fujitsu. We would like to ask Mr. Aikawa today about the background and purpose of his efforts as well as a future outlook. Thank you very much. Now, my first question is, and this I think is evolutionary. What is the background behind starting this project? System development, the issues there were becoming larger in scale and more complicated. These are the challenges in IT system development. As a result, it's taking more time to develop, and that is impacting delivery time as well as cost. Not just adverse circumstances. There's lack of IT engineers. That's another challenge. And engineer unit cost is skyrocketing. There are these many challenges in the field of IT system development. As we face all of these challenges, how can we overcome them? We've been thinking about this from before and trying to make effort, and ever since I started working, more than 20 years have passed, and agile development, there's been advancement there, but there's been no fundamental improvement in the development process. During the past two or three years, there's the advent of AI, including generative AI, and that is innovative, evolutionary, and that may fundamentally change the development process, and there's outstandingly high development speed, and that quality remains unchanged. That is the kind of environment that we can realize through AI, and originally, when existing systems were developed, we were consulting about the possibility of introducing the use of GenAI in using development, but at this, we discovered it could not lead to fundamental improvement, and we consulted with Fujitsu on this point, and Fujitsu kindly mentioned that we develop and work together. Thank you very much for the encouraging remarks. There are goals that we have to challenge and try to achieve and overcome. Source code and GenAI, there's a lot of advancement, announcements about them. Complete automation, though, is still a difficult challenge. And my next question. Through this initiative, I think we are in the midst of this initiative. How do you feel about co-creation with Fujitsu? First, when we first started to collaborate with Fujitsu, we were not sure how much progress we can make. But after we started to work with Fujitsu, we are now very much looking forward to the future. And what we did this time was we input the design, and we developed, and we conducted tests. And that was done automatically. And in the beginning, accuracy position was not so high, we have to admit. Through trial and error, we've been able to increase accuracy. There's been improvement in input, how the design is written. RAG development was also improved, and rule-based improvement was also made. LLM, large language model itself, is becoming more clever. This also enabled us to increase accuracy. Going forward as well, we want to continue to collaborate with Fujitsu. I think we can generate significant outcome. I think there's a lot of potential in this endeavor. In the future, various kinds of business work will utilize GenAI. We, as an operating company, want to use GenAI to define requirements and use that as input. In design, development, test, and release, we want to automate this entire process in an integrated manner. We are very much looking forward to this kind of a future. Thank you very much for indicating a very important point. The current business process can remain unchanged and automated. That cannot be done. We have to use AI. Designs have to be translated into the form that AI can read the design. The last question is, how do you expect the automation and efficiency of the existing system maintenance to change your business? If we're able to realize automation, the service that is offered to customers will undergo a sea change. Because we hope we can have this kind of a service. We hope to develop this kind of an IT system. There are various ideas in our minds. But because of many constraints of IT development costs and resources and return on investment, we have to be selective and prioritize the many ideas that we have. However, with the advent of GenAI, in an integrated manner, development can be possible, and then cost and time-related constraints can be eliminated. What we want to develop can be developed. I think we are entering into this kind of a world. By releasing various services to customers and by receiving customer feedback, we want to make further improvement. We want to run this cycle in a speedy manner. That is what we want to do in the future. If this virtuous cycle spreads within society, I think it will lead to a significant change. Fujitsu is working on cutting-edge technology. We want to continue to cooperate with each other so that we can transform society. Future system development and beyond that, agents will become the most important social infrastructure. I think we are making a very important step. For us, this is a kind of a future. In order to realize this future, let us continue to cooperate. We're very much looking forward to the future. Thank you very much, Mr. Aikawa and Uncertain (possibly Takahashi). If you could please return to your seats. Thank you. Hai. Aikawa-san, Uncertain (possibly Takahashi)-san, thank you very much. Now, as Fujitsu's SI is undergoing a transformation, Uvance is also exploring new value propositions. Today, we are pleased to introduce a new initiative powered by Uvance. This is the concept we have been promoting. Together with partners, companies who share common aspirations, we will combine technologies and expand solutions globally. That is powered by Uvance. When we established Uvance initially, we decided that we would not only focus on our own IPs, but we will integrate our IPs with the IPs of other partner companies. And in order to expand this ecosystem and pursue this concept, we established powered by Uvance, in which we combine the technologies for the solutions for our partners. For example, in the United States, city security geospatial recognition technology is a strong area of Aria. We are using video behavior analytics of Fujitsu Kozuchi, which is video analytics AI. This is used for prevention of crimes in urban areas. Regarding the solution, this is not only used for detection of hazards, but this can also be used for evacuation guidance during disasters and safe and secure urban development. To see how the collaboration between humans and AI will change the security field, please watch this video. The number one value that we're jointly going to provide with Fujitsu and Aria is proactively reducing crime. Specifically, there's large properties in Las Vegas. They have 5,000-10,000 cameras. Yet they only have five people looking at all these TVs. It's physically impossible to look at that. When crime happens, it's hurting somebody, obviously. Something's happening. If we can prevent that from happening, we win. The joint technologies will identify a pattern of behavior. If a fight is starting to happen, maybe we can deploy security guards quicker than no one else would be able to see. Fujitsu has strength in their vision AI technology. The accuracy is very high. We are competitive with others globally. For ARIA, they have their security expertise. Based on their security expertise, they can develop very valuable platforms for their end users, especially outside of Japan. The ARIA platform plugs in Fujitsu's technologies of EBA into a common area where it can be viewed and then disseminated to the right person at the right time to make action and decisions happen. Through AI geospatial recognition, through your current technologies with the video behavioral analytics, it's a perfect marriage. The culture is aligned. Society, communities, helping people. We believe in artificial intelligence. There's a lot of fear globally in the United States. If you look at it properly, it's here to help people. It's going to make people's lives better by doing this. I think mutually our companies believe these things. Fundamentally, that's our core of what we're at. Next is a case study of Fujitsu Kozuchi's skeletal recognition AI called Human Motion Analytics. Many of you may be aware of this because this is used as scoring aid for gymnastics competitions. It started with an intent to achieve fair scoring through the use of AI as a scoring aid, which recognizes the skeletal movement of humans. As announced recently, it has also been introduced to the National Training Center for Figure Skating. It accurately analyzes high-speed complex movements such as jumps and spins to support extension of capabilities of the athletes. So this is about human empowerment. Companies' DX and social challenges are other areas. But this focuses on extending human capabilities through the use of technology. This is an example that we use Uvance for. Please watch the video. Thank you. In order to do quadruple jumps, she wanted to know what she needs to improve. Compared to male skaters, she lacks muscular strength. So through AI recognition, she learned that she needs to improve her speed in jumping. Regarding this Human Motion Analytics, this is now extending beyond sports into the medical field. As announced today, Acer Medical in Taiwan has developed a solution to assess the risk of abnormalities in gait patterns. And this will be demonstrated in Taiwan. This is aimed at supporting early detection of various diseases such as dementia and Parkinson's disease, enabling early treatment. Now, we welcome Mr. Alan Lin, Chairman of Acer Medical from Taiwan, who will introduce their initiative. Alan, please join us on the stage. Dr. Lin, if you can please come on stage. Thank you very much. And please use the phones to listen to simultaneous interpretation. Good morning, everyone. Distinguished guests, this is my honor. Thank you today for me to present to you our new development with Fujitsu. I'm Dr. Alan Lin, the Chairman and CEO of Acer Medical. Acer Medical is a subsidiary company under the mother company Acer, which is a global computer brand. And Acer Medical focuses on AI applications in healthcare. What we're going to present to you is the Human Motion Analytics for early risk detection. What does that mean? If you have the disease, you are sitting in a neurologist's clinic. A doctor will tell you to do this walking test and measure the time up and go, sitting from sitting to walking out, turning, walking in, and turn around and sit again, but if you are affected by different types of diseases like Parkinsonism, epilepsy, frailty, stroke, the time in all these breakdowns will present differently, but it will be very difficult for a doctor to really measure it, so in the status quo, the current status in a clinical setting, the doctor will just use a stopwatch to observe manually, and sometimes there's a lack of standardization or lack of quantifiable output. Means the doctor might notice there's something wrong with the walking pattern, but they won't be able to quantify and measure it and put it in the medical record. And if they really want to make it very, very objective, they still need to use a sensor-based capture system, which could be very cumbersome and very difficult to scale up, so what we're presenting today is an AI solution that can automatically analyze the time that we just mentioned. And with this HMA, Human Motion Analytics technology, we can quantify all these indicators, and then with an EMR class system, so what we're trying to do is we want to move this AI in the application of gymnastics into health, so we are moving beyond sports into healthcare. Like in the system of gymnastics, the ref would need another pair of eyes. That's the AI to help the judgment. In the clinical setting as well, the doctor would need another pair of eyes, which is our AI, to help make precision, to increase the precision of diagnosis. So indeed, let's dive into it in detail. You can take a video using your smartphone. And the video will go into the analytics, the system, the skeleton system, recognition system. And with the analysis of the system, it can give you an output that's actually medically relevant to assist a doctor in making precise decisions. So let's look into a more deeper way. But I will give you a story. We took a normal person and we told them to walk and then upload the video in the system to measure the move of the angle of their joint, left knee and right knee. And this experimental group is a patient with sickness. And we noticed that the sick patient has a very unstable knee when they move. So this angle is like very, very wavy. But we realized that the right knee of the normal patient, normal person that we picked, also has some instability, which is so subtle that even the doctor or the family didn't notice. But the AI system can capture it. And the right-hand side, the central mass is also presenting the deviation of central mass of a dementia patient when they walk. So the system will be able to summarize all these measurements into this radar chart. What does this do? This can actually help monitor the progression of the disease when you actually go through rehabilitation. Actually, these different metrics can improve through time. For example, this patient, they have a lower score in cadence, which is the flow or the rhythm of their walk. But through time, it could improve after intervention. So that creates unlimited imagination for our clinical application. For example, for orthopedics, you can use this AI system to monitor the post-operation and pre-operation improvement. Or in a care center, you can monitor the patient that's hospitalized. Or in rehabilitation, you can actually monitor the progression of the improvement of the patient. So what we are trying to say today is that by collaborating with Fujitsu, Acer Medical is trying to implement this in a clinical setting so that together we can make people healthier. But not only living longer, but live healthier, but having a healthier life longer. So that we are creating cross-industry AI collaboration for healthy longevity. Thank you. Alan, thank you so much. I'm back on the stage. Yes. So before we continue our dialogue, I just want to kind of invite Oiwa-san, who is actually the project leader. Mr. Oiwa, responsible for the project, will be coming on the stage. Question to you. What did you think when you first saw this Human Motion Analytics? You thought it was kind of usable for your proposition or solution? Yes. When I first saw the system with the application for gymnastics and the referee system, and because it will give so much detail and precise measurement of the movement of angle, and that made us think of the application in healthcare. Because in healthcare, some measurement is based on very, very empirical. Like if a doctor or neurologist is trained in a very sophisticated way, they can tell the difference that they couldn't really articulate or measure the difference of the gait. But with this system, actually, that will give us all these quantifiable metrics that will help them actually put those metrics in the structured medical record. And that can actually make the doctors actually talk to each other in a more objective way. OK, that's great. So just like a common behavior pattern for diagnosis, it's very important, right, to kind of define. Yes, yes, yes. OK, thank you so much. Thank you. Then I'm going to ask Oiwa-san. So I think not only in Japan, Asia, across the world, we have an aging society. So as a project leader, what do you think the solution, the joint solution, is going to, what type of impact do you think it will bring to the society as a whole? As the global population ages, we are facing not only the medical challenges, but also social and economic issues. And this initiative offers the meaningful value to the society by supporting healthier living for all. It helps people maintain their quality of life while also easing practical and emotional burden on families. It contributes to lowering medical and social security costs as well. So this is not just a solution for doctors and patients, but it is a contribution to building a more resilient and health-conscious society. Thank you so much. I think this quality of life for everybody is like a very important concept through technology, right, Alan-san? So thank you so much, Mr. Alan, Oiwa-san. I hope that this solution is going to bring a big, big impact to the world. Thank you so much. Thank you. Thank you very much. Thank you very much. Thank you. Dr. Alan and Oiwa-san, thank you very much. Yes. Now, so far, you have seen various case studies of events, starting from supply chain optimization, and as AI is introduced to estimate P&L impact by contingencies and tariffs. That leads to behavioral transformation. In systems development and maintenance, rationalization is carried out, and also creating safe and secure communities and expanding human capabilities of athletes. People and AI work together in every case study. Fujitsu Uvance aims to leverage data and AI to solve both business and social challenges. Adding AI agents here further enhances the quality and speed of decision-making. Furthermore, what is important is we will have multiple agents collaborating across operations and departments, which enables total optimization. In the future, agents will be able to collaborate not only within companies, but also between companies to do sophisticated decision-making across industries and ecosystems. This is our vision. This is what Fujitsu Uvance envisions as cross-industry problem-solving. Independent AI agents of different companies will be linked together to collaborate to solve various challenges. Once again, as a recap, Fujitsu's AI agent consists of three pillars. First is business-specific agents. For many years, together with customers, we have been developing systems. And we have expertise in manufacturing, logistics. AI agents, in order for them to replace the work, they need to understand the business as well as we're better than humans. That is how we will be preparing our AI agents based on what we have learned with our customers over many years in the past. Another important aspect is technology. We can utilize a Japanese language-enhanced LLM to accurately interpret corporate data. And we need to do structuring of data necessary for AI agents to maximize their capabilities. We can also develop and maintain a foundation for understanding the context. We believe that we, as a technology company with a deep understanding of the field, can accomplish this. The second is the promotion of multi-agent and multi-vendor concept. For performing complex tasks, we need to collaborate with SAP, ServiceNow, and Salesforce, for example, with multi-agent collaboration, where agents specializing in different fields of expertise collaborate. Together with Salesforce, SAP, ServiceNow, and Microsoft, we are establishing a mechanism to work with their agents. To give a concrete example, if you want to update the business meeting information in CRM, after meeting with a customer, simply make a request via Teams chat to read the business meeting details from the minutes. The Microsoft AI Agent Copilot and Salesforce's Agentforce work together to automatically update the meeting record. We also have future visions. For better transaction and conversion rate, there's much more we will be able to do. Another aspect is supply chain, so based on a demand forecasting function of Fujitsu, a supply chain agent may decide to run a promotional campaign to eliminate excess inventory, and then the best Salesforce agent for the task is selected, and a request and content will be generated. These will work in a collaborative format, and we also have GK Software as a company, which can be combined with the dynamic pricing concept. Different aspects in different industries will be connected eventually. In this way, what is important is rather than collecting everything on a single platform, agents specialized for each system work together in the right places and are connected in a network, and there are various AI agents developed by different companies. The third important point is reliability, how we ensure reliability and appropriate governance. A guardian agent is a concept that has been applied around the world, and this is going to be critical. In utilizing AI agents, there is an issue of prompt injection. What kind of countermeasures we can take against hallucination is an important matter. And together with AI Act, we can handle unmasked data to connect with other data. And also in the area of AI ethics, we have been researching for over 10 years now. So based on AI ethics in the area of security and privacy, we have accumulated know-how. So together with a reduction of hallucination, we believe we'll be able to achieve appropriate governance. Centered around our lab and based on our expertise of Guardian agents, we would like to develop further technologies. Safe and secure AI and solutions with such AI are going to be our important mission. Last but not least, AI agents will be connected across different industries. Today, we have presented the evolution of Uvance over the past year, which included diverse case studies. But what was common among them was Uvance, which is the foundation of Uvance. We are trying to move forward to solve social challenges. This cannot be done alone by ourselves, but we are trying to work together with partner companies. So under the umbrella of the Uvance concept, we would like to work with companies, public sectors, and we would also like to focus on extending human capabilities. That is how we are envisioning to evolve Uvance. We are going to be having demonstrations afterwards, so you will be able to give us your feedback after looking at those demonstrations. Thank you very much for joining us once again.