Customer Case Study: Fujitsu Composite AI and Semantic Kernel

Matthew Bolanos

Japanese multinational Fujitsu, a pioneer of information and communications technology, has been transforming industries with innovative solutions since 1935. With a workforce of 124,000 dedicated professionals across 50 countries, Fujitsu is committed to building trust and fostering sustainability through its groundbreaking technologies.

A diverse portfolio that includes everything from IT services to server equipment has a new member: AI. Fujitsu’s AI solutions (branded as Fujitsu Kozuchi) are broken into seven areas:

• Generative AI
• AutoML
• Predictive Analysis
• Text
• Vision
• AI Trust
• XAI

With the help of Semantic Kernel, we’ve been able to stack these technologies together to better solve customer needs from a single platform.

A new frontier: Fujitsu Composite AI

Fujitsu Composite AI is a unique combination of AI technologies that can understand abstract business problems through chat-style dialogue. It automatically analyzes a situation, searching for and proposing specific solutions based on past data.

It dramatically expands the application and use cases of AI by combining multiple technologies to build the ideal solution. Complex issues that a single AI model could not address can now be solved quickly and at a low cost.

From ambiguous to automatic with Semantic Kernel

Semantic Kernel is an SDK that, as the documentation says, lets you “actually do something productive.” By using it to connect Composite AI component technologies, we can address real customer needs.

Our pipeline breaks down ambiguous instructions and automatically combines multiple AIs to create an advanced model capable of delivering a solution. In fact, if the required model doesn’t exist, one that is bespoke and solution-fit will be generated.

Real problems, real solutions: Composite AI case studies

Fujitsu Composite AI is already being applied to real customer data, creating efficiencies and solutions for issues that were previously cumbersome or resource-intensive.

Nakayama Transportation

Composite AI powers Nakayama Transportation’s automated vehicle dispatch system. The system analyzes the driving and restricted time of the drivers, generating an efficient plan while complying with laws and regulations.

Nakayama Unyu has given the AI solution high praise for its ability to manage both vehicle dispatching and working hours in a single tool.

Results: The time it takes to create a dispatch plan has dropped from several hours to 10 minutes.

  Fujitsu Customer Support

Internally, we’ve used Composite AI to accurately predict customer support requirements and optimize resource allocation. The platform analyzes incident management logs, predicts the future of any incident (how many days before it is resolved), and suggests staffing allocation.

Results: The new incident management system is 25% more efficient than the previous system.

Thoughts from the Semantic Kernel team

Semantic Kernel’s PM Matthew Bolaños had this to say about the Composite AI integration:

“I was very impressed with the solution that Fujitsu has implemented. It’s great to see how they have been able to use Semantic Kernel to improve the customer experience. It’s positive to see they’ve leveraged Semantic Kernel to integrate their AI technology as Composite AI and apply it to several real business fields.”

Summary

By enhancing Fujitsu AI technologies with Semantic Kernel, we can design a flexible solution pipeline and solve customers’ real problems. Composite AI automatically combines the most appropriate AI tech for any given task. In the future, we plan not only to deepen our collaboration with Semantic Kernel but also to explore integrating with Microsoft Fabric as a data source for Composite AI. We believe this integration has the potential to greatly enhance our capabilities and provide even more value to our customers.

We’ve only brushed the surface of its orchestration capabilities. Their importance will only grow as we continue to explore, innovate, and use these features more extensively.

References

Learn more about Fujitsu , the Fujitsu AI (Kozuchi) platform and Composite AI. For a more in-depth look, check out the whitepaper for Composite AI.

Learn more about Semantic Kernel.

Matthew Bolanos

Follow