Fujitsu Ltd.
Fujitsu Ltd.
09/05/2024 | Press release | Distributed by Public on 09/04/2024 18:09
Fujitsu and MoBagel revolutionize business processes through accelerated AI prediction
Fujitsu and MoBagel revolutionize business processes through accelerated AI prediction
Fujitsu Limited, MoBagel Inc.
Kawasaki and California, September 5, 2024
Fujitsu Limited and MoBagel Inc. today announced that Fujitsu has provided AI technology from its Fujitsu Kozuchi service for use in Decanter AI, an AutoML platform developed by US-based AI startup MoBagel Inc. This new solution, designed to accelerate AI-driven predictions, will be available globally starting from September 5, 2024 and is born out of a strategic partnership between the two companies established in December 2023.
Fujitsu Kozuchi AutoML, which automates the construction of machine learning models, and Wide Learning™, a part of the explainable AI technology Fujitsu Kozuchi XAI, are being provided to MoBagel under license. By integrating Fujitsu's AI services into Mobagel's Decanter AI platform, the two companies have enabled the creation of advanced AI models that are both accurate and cost-effective, while streamlining data analysis. With this solution, the two companies will drive AI innovation and help more businesses and organizations achieve faster AI-driven transformation.
How the technology speeds up AI predictions
The AI prediction acceleration solution integrated into Decanter AI includes three key features.
(1) Data Verification and Anomaly Scoring:
The input data are validated using Decanter AI's processing technology to ensure accurate and efficient modeling. This scoring system flags the data as being either "Healthy", "Moderate", or "Critical" based on its integrity and relevance. This system ensures that the dataset is accurately evaluated for optimal use in analysis and modeling and will give suggestions on how to refine the data for better management.
The input data are validated using Decanter AI's processing technology to ensure accurate and efficient modeling. This scoring system flags the data as being either "Healthy", "Moderate", or "Critical" based on its integrity and relevance. This system ensures that the dataset is accurately evaluated for optimal use in analysis and modeling and will give suggestions on how to refine the data for better management.
(2) Algorithm Recommendation System:
The Algorithm Recommendation System based on Fujitsu Kozuchi AutoML technology suggests the most suitable machine learning models. This improves modeling efficiency and accuracy.
The Algorithm Recommendation System based on Fujitsu Kozuchi AutoML technology suggests the most suitable machine learning models. This improves modeling efficiency and accuracy.
(3) Deriving Actionable Insights through Feature Analysis:
After the data are analyzed by the chosen machine learning models, Feature Analysis based on Wide Learning™ identifies key features and illustrates importance and relevance using weighted combinations. The identified features are then interpreted and summarized by a large language model and reported in natural language.
The figure below illustrates Decanter AI being used to identify people who are likely to purchase car insurance. The data verification and anomaly scoring stage checks past purchase data for outliers and eliminates anomalous entries. The algorithm recommendation system suggests the most accurate model for predictions, in this case the random forest model. Finally, Feature Analysis technology ascertains the most important features for identifying individuals likely to purchase car insurance and adds weighted values. This analysis is then translated into natural language through a language learning model.
After the data are analyzed by the chosen machine learning models, Feature Analysis based on Wide Learning™ identifies key features and illustrates importance and relevance using weighted combinations. The identified features are then interpreted and summarized by a large language model and reported in natural language.
The figure below illustrates Decanter AI being used to identify people who are likely to purchase car insurance. The data verification and anomaly scoring stage checks past purchase data for outliers and eliminates anomalous entries. The algorithm recommendation system suggests the most accurate model for predictions, in this case the random forest model. Finally, Feature Analysis technology ascertains the most important features for identifying individuals likely to purchase car insurance and adds weighted values. This analysis is then translated into natural language through a language learning model.
Figure 1: An example of Decanter AI being used to predict car insurance purchasing behavior
The incorporation of Fujitsu Kozuchi AutoML and Wide Learning™ into Decanter AI, enables the system to recommend an optimized machine learning model four times faster than was previously possible while maintaining accuracy. In addition, feature analysis powered by Wide Learning™ technology generates actionable recommendations by making relationships easier to find through its use of natural language. This AI solution is highly adaptable, cost-effective, and can be deployed quickly.
Hideto (Ted) Okada, SVP and Head of Technology Strategy Unit, Fujitsu Limited, comments:
"Fujitsu's suite of advanced AI services provided through Fujitsu Kozuchi make it a leading player in the enterprise AI market . By providing Fujitsu's expertise to innovative startups such as MoBagel we are helping our customers to rapidly adopt AI and contributing to the development of new AI offerings."
Adms Chung, CEO of MoBagel Inc., comments:
"We are thrilled to partner with Fujitsu, as together we have introduced innovative AI technology that addresses the industry's most critical challenges: accuracy, speed, and cost-efficiency in AI transformation. By integrating Fujitsu AutoML and Wide Learning into MoBagel's AI platform, we improve the development and integration of AI models, making them more explainable and better aligned with business impacts. This collaboration enables businesses to create AI applications or AI agents faster, more reliably, and cost-effectively, bringing us closer to our vision of building AI together and beyond."
Future Plans
MoBagel, in collaboration with the team at Deloitte Touche Tohmatsu Limited, will conduct a PoC for utilizing this solution in financial crime prevention. The PoC will take the form of a hands-on workshop where participants use Decanter AI to analyze large volumes of trading data and rapidly identify unusual behavior and suspicious trading patterns.
Fujitsu and MoBagel will continue to deliver AI solutions to customers across industries empowering them to transform and streamline their business processes.
About Fujitsu
Fujitsu's purpose is to make the world more sustainable by building trust in society through innovation. As the digital transformation partner of choice for customers in over 100 countries, our 124,000 employees work to resolve some of the greatest challenges facing humanity. Our range of services and solutions draw on five key technologies: Computing, Networks, AI, Data & Security, and Converging Technologies, which we bring together to deliver sustainability transformation. Fujitsu Limited (TSE:6702) reported consolidated revenues of 3.7 trillion yen (US$26 billion) for the fiscal year ended March 31, 2024 and remains the top digital services company in Japan by market share. Find out more: www.fujitsu.com.
About MoBagel
MoBagel is a Gartner-recognized AI/ML platform committed to empowering users to make informed, data-driven decisions seamlessly. Headquartered in Silicon Valley with seven global offices including Tokyo, Taipei, and Singapore, MoBagel has assisted over 11,000 brands in deploying AI to create a more resilient future.
Using generative AI and machine learning technologies, we develop multiple AI Agents that enhance business process automation and foster precise decision-making in critical areas such as sales & marketing, supply chain, finance, and manufacturing. Our platform significantly reduces the typical AI implementation timeline from 12 months to just 1 month, enabling businesses to rapidly address challenges through swift model creation, testing, and deployment.
Press Contacts
Fujitsu Limited
Public and Investor Relations Division
Inquiries
MoBagel Inc.
E-mail: [email protected]
All company or product names mentioned herein are trademarks or registered trademarks of their respective owners. Information provided in this press release is accurate at time of publication and is subject to change without advance notice.
Date: 5 September, 2024
City: Kawasaki and California
Company: Fujitsu Limited, MoBagel Inc.