Getting real results from your AI transformation
How AI can transform companies that struggle to change
Are you really getting the results you expect from AI?
In many cases, the answer is no. A recent, authorative report by MIT, for example, shows that 95% of organizations using generative AI have yet to see profits.
Too often, discussions focus on trends or quick wins, without a full view of AI initiatives. The result? Companies achieve only small efficiency gains with little lasting impact. To unlock AI’s full potential, businesses need to rethink processes and data architectures, shaping them to deliver real, measurable value.
Uvance Wayfinders, Fujitsu’s consulting arm, helps organizations uncover their true strengths in processes, data, and expertise, and design transformations that build on them. By combining this insight with years of delivery experience, Wayfinders goes beyond strategy and recommendations, supporting the full implementation of enterprise transformations for the AI era.
What sets Wayfinders apart is its ability to leverage proprietary advanced technologies developed by Fujitsu Research ahead of the competition, while deploying consultants alongside AI-savvy engineers and architects in a forward-deployed model, to solve challenges end-to-end.
* This article is an excerpt from an advertisement published on Nikkei Business Online with permission from Nikkei BP. (Unauthorized reproduction prohibited)
* Titles and affiliations are as of the time of the interview.
"Do you really need that task?" The need for business process transformation
While AI is a powerful tool, it is not a cure-all. Satoshi Mihara, who leads technology consulting at Wayfinders, puts it simply: "Depending on where and how AI is applied, there are many cases where expected results are not achieved."
A common reason is trying to implement AI without adapting existing workflows and systems, which were originally designed around human workers. So, what perspectives are essential for successfully driving enterprise transformation with AI? Mihara highlights three key viewpoints:
The first perspective is the need to transform business processes and IT simultaneously. Existing workflows and systems must be reshaped into “AI-ready” forms that make it easier to apply AI effectively. "Start by simplifying each task so AI can handle it more easily," explains Mihara. "The IT world has many methods for simplification, and these should be applied to business processes as well. Introducing AI only partially has limited impact. Workflows need to be redesigned end-to-end, integrating AI throughout while clearly defining the roles of humans and AI. Systems must also be transformed to process business events in real time, in a truly data-driven way."
The second perspective is to take an integrated approach to data and AI, rather than focusing on AI alone. After all, without data, AI can achieve nothing. A key factor is how well a company’s data assets are organized into forms that AI can easily use. Traditional data systems often store information with a one-day delay, but agent-based AI requires real-time access. Additionally, just as humans need to understand and interpret data, AI must also be able to “know” what the data means.
The third perspective is strengthening and evolving security. Rather than continuing with traditional approaches based primarily on checklist-based assessments, companies need to build advanced security systems that prepare for AI-driven attacks and utilize AI on the defensive side as well.
Hideto Okada, Head of AI Strategy and Business Development Unit, points out: "Many companies are beginning to introduce AI, but not many have yet achieved 'revolutionary productivity improvements.' The reason is clear. AI cannot demonstrate true effectiveness by simply automating complex, fragmented business processes as they are. To genuinely enhance productivity, it's essential to simplify, standardize, and redesign the workflows themselves before introducing AI."
When companies introduce AI, they often begin with simple tools, such as chatbots, to support everyday tasks, and then gradually extend AI into parts of existing workflows. For example, automatically generating program source code can improve operational efficiency. While more companies are reaching this stage, the benefits are often limited to partial efficiency gains. If humans must verify AI outputs at every step, scaling its effectiveness becomes difficult.
"Simply inserting AI into parts of current workflows without changing them limits effectiveness because the workflows remain unorganized. If humans have to check AI results at every turn, costs may actually increase," says Okada.
What's crucial is the next stage: rebuilding workflows to be AI-driven. This means simplifying, standardizing, and redesigning workflows around AI, as mentioned earlier. "A common example is report creation. We often hear about replacing manual or Excel-based tasks with AI. But the question is whether the report creation itself is even necessary in the first place. It's important to review the entire workflow from this perspective and simplify it," explains Okada.
Such reviews cannot be conducted by IT department members alone. The return on AI investment depends on business and IT departments collaborating to fundamentally reconsider business processes with AI utilization as a prerequisite.
When data gets stale, where is the ‘secret sauce’?
It’s well known that aligning data formats and ensuring quality are key to making AI effective. Without integrated data, AI can fail. But that’s not enough. "We need to examine how data is held and processed from the temporal perspective of freshness," says Mihara.
In many organizations, business data is stored in data warehouses or lakes, which AI then references. If this is mainly batch-based, the AI may be working with data that is hours or even a full day old. When AI relies on outdated information, its outputs can quickly become misaligned with reality.Data freshness is struggling to keep up with the speed AI requires. To fully leverage AI, companies must go beyond preparing data for human analysis and rethink how data is collected, updated, and processed in real time for AI operations.
Among these, leveraging data that connects to a company's strengths as assets has become particularly important. Seishi Okamoto, Director of Fujitsu Research, states: "With the arrival of the AI era, the time has come for companies to identify which data among their vast holdings is the source of competitive advantage and to utilize it." Okamoto refers to such data collections as the "secret sauce" that generates a company's competitiveness. If this "secret sauce" can be utilized with AI, it can lead to significant market competitive advantage.
How can a company’s unique “secret sauce” be incorporated into AI? Fujitsu Research has been working on a solution. "A company’s ‘secret sauce’ often comes from unstructured operational data, such as design drawings or electronic medical records," explains Okamoto. "Fujitsu Research has developed technology that structures data that computers traditionally struggled to handle, allowing AI to understand and make use of it effectively."
Wayfinders helps companies unlock the value of their unique operational data by transforming it into AI-ready formats, using appropriate technologies, and tailoring the approach to each customer to ensure successful AI adoption.
Furthermore, Wayfinders adds value in uncovering a company’s true “secret sauce.” A company’s real strengths may lie in its processes, its data, or the way work is executed, including the use of AI. Companies may assume they already know the answer, but a thorough, multifaceted assessment is essential to identify where the genuine advantages lie.
Mihara explains: "It’s difficult for companies to do this thoroughly on their own. From an expert standpoint, we examine multiple angles and objectively identify where the real strengths lie."
Wayfinders supports organizations from defining their strengths to implementing advanced technologies, going beyond optimization within a single company to include comparative insights from competitors and related industries.
The foundation of Wayfinders’ approach is Fujitsu’s uniqueness. In business environments where agent-based AI operates, any system downtime translates directly into business downtime. Years of building highly reliable systems for critical social infrastructure have established a delivery capability few can match, while Fujitsu’s in-house development of technologies in areas others have not explored adds further advantage. On top of this, consulting expertise is layered to guide strategy and implementation.
Mihara emphasizes: "We're not competing solely on consulting. Within a single project, we can advance strategy, technology application, and delivery simultaneously, solving challenges comprehensively. That’s Wayfinders’ strength."
Leveraging proprietary technology: Fujitsu's consulting approach
Fujitsu is also transforming its own operations through the use of AI. “The wave of generative AI is now reaching system development that underpins social infrastructure and large-scale enterprise operations,” says Okada. While AI tools that automate source-code generation and testing are already in use, Fujitsu is taking the next step beyond these capabilities.
One pressing challenge lies in the modernization of long-standing core business systems. Many Japanese companies continue to rely on large-scale systems that have been in place for decades. Updating them requires significant investment and effort, and carries the risk of disrupting critical day-to-day operations. As a result, executive leadership often faces a difficult decision: migrate immediately to modern systems, or extend the life of existing ones for several more years.
When companies choose the latter, hidden costs quickly accumulate. Even minor changes—such as updates to laws and regulations, new product launches, or service specification changes—require manual modifications to legacy systems. Each adjustment adds further time, cost, and operational burden.
Fujitsu Research is applying proprietary technologies to enable end-to-end automation through AI. For example, when a user inputs instructions such as “this regulation has changed in the following way” into a generative AI prompt, the AI compares the old and new regulations, identifies which system assets must be updated, and determines how those changes should be made. The entire workflow—from implementing the modifications to testing—can then be automated. What was once limited to partial automation or experimental prototypes has now evolved into a fully automated process, capable of delivering production-ready systems simply by specifying the desired outcome.
As the adoption of generative AI accelerates, broader societal challenges have also emerged, including rising electricity and water consumption. To address these issues, Fujitsu Research is developing advanced efficiency-focused technologies. These include “1-bit quantization,” which improves AI processing efficiency while significantly reducing power consumption, and “specialized AI distillation,” which compresses domain-specific knowledge to optimize AI model architectures.
In internal validation using proprietary data, specialized AI distillation delivered substantial gains: GPU memory requirements and operational costs were each reduced by 70%, inference speed increased by a factor of 11, and accuracy improved by 43%.
These more sustainable technologies have already matured to the point where they can be delivered to customers. Wayfinders consulting enables enterprise transformation in the AI era by applying these proprietary innovations developed by Fujitsu Research.
The evolution of security – essential for business continuity
“In the AI era, strengthening and evolving security is no longer optional. It is a management issue that directly affects business continuity and must be addressed with urgency,” says Mihara. As AI is increasingly used not only to defend systems but also to power cyberattacks, traditional single-layer security measures are no longer sufficient.
This shift is also reshaping the security landscape. Fujitsu Research is advancing defensive security through AI-driven research and development. Wayfinders’ white-hat hackers conduct simulated attacks from the perspective of real attackers to identify vulnerabilities. The insights and data gained from these exercises are then used to design more advanced, AI-enabled defense systems.
“Security is also a key focus of our work on multi-AI agent technology,” adds Okamoto. “We have developed systems in which multiple AI agents are assigned to offensive and defensive roles. These agents continuously compete, learn from each other’s actions, and become more sophisticated over time. Other AI agents objectively monitor and evaluate the situation, enabling multiple agents to collaborate in solving complex security challenges.”
Through Wayfinders consulting, Fujitsu aims to bring these advanced security technologies into real-world business environments and broader society, helping organizations strengthen resilience in the AI era.
Okada adds, “It’s also essential to ensure that AI itself is used safely and securely. When multiple agents interact, independent monitoring agents are needed to confirm they operate within established rules.” Without proper governance, AI could become a new management risk. This means that, beyond technology, rule-making and institutional oversight are critical. Fujitsu participates in the AI Safety Institute (AISI) in Japan, established with the Cabinet Office and relevant ministries of Japan, to study AI safety from legal and societal perspectives.
Mihara emphasizes the importance of a strategic approach: “To turn AI and data into true sources of competitive advantage—rather than following trends or engaging in superficial discussions—management must clearly define the overall framework of initiatives and drive them forward with strong, top-down leadership.” Supporting companies in this effort, by combining broad perspective, deep expertise, and Fujitsu’s advanced technologies, remains the core mission of Wayfinders.
Satoshi Mihara
Head of Global Technology Practice, Managing Partner, Uvance Wayfinders
Head of Global Technology Practice, Managing Partner, Uvance Wayfinders, Fujitsu
Satoshi Mihara graduated from the School of Political Science and Economics at Waseda University. After working at a domestic systems integrator, an overseas IT vendor, and in business management within the retail industry, he joined Accenture in 2007. Subsequently, he led data utilization projects as Managing Director and engaged in consulting primarily for the financial and retail industries. He joined Fujitsu in June 2024. Leveraging his extensive technical knowledge and experience, he is driving global expansion and transformation in the technology consulting domain at Uvance Wayfinders.
Hideto Okada
SVP, Head of AI Strategy and Business Development Unit
Hideto Okada joined Fujitsu in 1991. After working on local government business application packages, he promoted "Municipal DX" as General Manager of the Second Public Administration Solutions Business Unit from 2016. In 2020, he transferred to Fujitsu Laboratories, primarily taking on the role of strengthening collaboration with business divisions. From 2021, as Head of Technology Strategy, he was responsible for company-wide technology strategy and experienced assignment to Silicon Valley. He assumed his current position in April 2025.
Seishi Okamoto
Executive Officer, Senior Executive Vice President, Director of Fujitsu Research, Doctor of Science
Seishi Okamoto joined Fujitsu Laboratories in 1991 and has been engaged in research and development of artificial intelligence including machine learning, inference, natural language processing, and knowledge search. For three years from 2011, at Fujitsu, he was engaged in big data new business development and data scientist training. After serving as Director of the Artificial Intelligence Research Center, Director of the Artificial Intelligence Laboratory, and Fellow at Fujitsu Laboratories, he became Executive Officer EVP and Director of Fujitsu Research in April 2023. He assumed his current position in April 2025. He also serves concurrently as Visiting Professor at Tokyo Medical and Dental University.
Contact us