KIBIT, AI Engine of FRONTEO
Application domains of KIBIT expanding
Utilizing KIBIT, FRONTEO applied a “human-like language processing” beginning with legal domains and has come to expand the application into three areas of business intelligence (finance, intellectual property and human resources domains, etc.), healthcare (medical and pharmaceutical domains, etc.) and digital communications (robot domains etc.), successfully introducing KIBIT to various on-site scenes in diversified domains. KIBIT also has the advantage of being installed at relatively low costs, as it is offered in an application format.
KIBIT in evolution
At present, KIBIT is incorporated in FRONTEO’s applications to operate as an “AI engine.” In next-generation models, however, KIBIT will be separated from applications, evolving into a “KIBIT platform” that operates autonomously in response to requests from outside.
Research is under way to equip FRONTEO with multiple modalities, a mechanism that allows simultaneous processing of different types of data like texts, images and voices, to make it possible to analyze data other than language. Moreover, FRONTEO is working on the idea of creating an autonomous AI platform, in which multiple AI machines communicate with each other in an autonomous and distributed manner to process data so that the scope of analysis expands automatically at a high speed.
The future KIBIT platform will serve as the integrated system supporting these complicated structures and autonomously processing big data to generate high-level, integrated results.
The future KIBIT will bring
Presently, when introducing AI, it is typical that data scientists or engineers specialized in machine learning work to build systems and incorporate AI machines. This imposes a large cost on the customers introducing AI, making it the case that not everybody can easily utilize AI.
To address this issue, FRONTEO is striving to develop a KIBIT platform that will enable anyone to readily use the know-how possessed by individual data scientists or machine learning engineers so that the benefits of KIBIT should be provided to a larger number of people in a shorter period of time.
At FRONTEO, we believe that the progress in AI is largely divided into two stages. The first is the stage in which AI and data will be increasingly used within existing organizations to drastically improve productivity. For example, AI would more efficiently analyze customers’ voices at call centers, and AI for conducting production management would be introduced at plants to enable on-demand production.
The next is the stage in which AI machines will coordinate with each other beyond the boundaries of existing organizations, promoting use of data on an even wider scope. For example, AI at a call center determining that a certain new product is provoking many complaint calls would coordinate with the AI system in charge of production management at the related plant to temporarily reduce the production volume of the product.
Moreover, it is expected that such coordination between AI machines will be conducted not only among existing organizations of an industry but beyond the boundaries of industries to create complicated eco-systems. Over the course of this process, it will become popular for AI to analyze multiple modalities including language, figures and images in an integrated manner.
Going forward, with KIBIT making progress to realize the next-generation “KIBIT platform,” it will become possible to conduct more advanced and easier analysis in response to modalities other than language and through mutual learning by AI machines. Stay tuned for the next-generation of KIBIT that will drive and shape a new future.