On September 16th, Fujitsu will develop an AI scoring platform service that AI makes scoring predictions and presents the next effective action, and supports operations in the financial industry such as sales marketing, loan screening, fraud detection, etc. “FUJITSU Finplex AI Scoring Platform Service EnsemBiz (Finplex EnsemBiz)” is now available.
In the financial industry, promotion of DX using the latest technologies such as Fintech and AI is indispensable for improving competitiveness, and active utilization is being called for. On the other hand, in financial services that require reliability and convenience, it takes time to put into actual operation. Is required. In particular, the fact that AI does not know the basis of the results derived is one of the factors that hinder the use of AI for management and investment decisions, and the shortage of AI human resources and the preparation of a large amount of learning data are issues.
The new service incorporates the AI technology “Wide Learning” developed by Fujitsu Laboratories, enabling highly accurate predictions with a small amount of data and the presentation of the next action. For example, an efficient approach in sales marketing activities. , It can be used for various operations such as improving the efficiency of loan examination operations and avoiding risks by detecting fraudulent claims and fraud.
In addition, by enumerating and analyzing a comprehensive combination of data, predictions can be made, and by presenting the combination of data that has a strong influence on the prediction results as a basis, the results can be understood by humans and are highly convincing. Can be obtained.
Furthermore, in the past AI utilization, AI automatically proposes multiple actions to be taken next, which were analyzed by data scientists based on the prediction results, puts action items in the data used for prediction, and analyzes the combination and results. By doing so, the next action can be presented. This will enable the selection and implementation of effective and efficient approaches, support business and business decision-making, and automate analysis to reduce the burden of business operations.
In addition, data scientists generally perform tasks such as data cleansing, data processing, and mass parameter tuning that are performed when utilizing AI, but data can be easily preprocessed by GUI operation, and the BI tool “Tableau” By linking with “, the prediction results can be displayed as a graph, so it can be used easily without specialized knowledge.
Since valid data is selected and predicted from all combinations of data, it is possible to derive features in the data even with a small amount of data, so there is no need to prepare a large amount of data for AI learning, and it is new. It is possible to apply AI to the area.
There are two types of provision, on-premise type by package and SaaS type, and implementation support and demonstration introduction are implemented. The basic package license starts at 600,000 yen per month, and sales are planned to be 1 billion yen by the end of 2023.