Fujitsu Composite AI whitepaper published

May 22, 2024


In recent years, the evolution of AI technology, particularly generative AI, has enabled its application to various business challenges. However, a single AI model often cannot fully address all business issues. To tackle this problem, an approach called "Composite AI," which combines multiple AIs, is gaining lots of attentions. This Composite AI is expected to efficiently solve business challenges by breaking them down into smaller parts and using specialized AIs for each. However, achieving this requires knowledge in both technology and business, posing high barriers.

Fujitsu has released a white paper on "Fujitsu Composite AI," which understands abstract business challenges through chat-based interactions and automatically explores and proposes concrete solutions using the optimal AI models and data from countless possibilities. This technology is designed to effectively address complex issues that are difficult for traditional single AI models to handle by combining multiple AIs, thereby significantly expanding the scope of AI utilization. With this advancement, AI technology can be provided more cost-effectively and quickly to areas that have been difficult to reach.

The "Fujitsu Composite AI" introduced in this white paper was developed in collaboration with Microsoft's AI framework, "Semantic Kernel". The results were showcased in a breakout session* at MS Build held from May 21, 2024, so please take a look at that as well.

Inquiries regarding this matter


  • facebook
  • twitter
  • LinkedIn
  • LINE
  • Weibo



Connect with Fujitsu Research

Top of Page