[FinTech] AI–based solution for automating data processing –Launch of initiatives to improve productivity within the financial industry in cooperation with regional financial institutions

Mizuho Financial Group, Inc. (President & Group CEO: Tatsufumi Sakai), Mizuho Bank, Ltd. (President & CEO: Koji Fujiwara), Mizuho Information & Research Institute, Inc. (President & CEO: Junichi Nishizawa), and Blue Lab, Co., Ltd. (President, CEO, & Representative Director: Daisuke Yamada) have started to implement digital innovation initiatives in non–competitive areas throughout the financial sector for the purpose of providing regional financial institutions with an AI–based solution for automating data processing of hand–written documents and non–standard forms, using artificial intelligence (AI), optical character recognition (OCR), and robotic process automation (RPA) technology. This solution has been dubbed "The AOR™".

Specifically, we have started to formulate plans for the seamless introduction of this solution in operational systems at financial institutions. The plan is being formulated by conducting proof–of–concept testing on The AOR™ in cooperation with six regional financial institutions and NTT Data Corporation, which operates joint operations centers for financial institutions, and by working collaboratively with two companies engaged in providing services related to back office workflow systems. We aim to substantially improve the productivity of common operational functions in the financial industry by establishing a framework for providing this solution to regional financial institutions.

In addition, Mizuho Bank has invested in Cogent Labs Inc. (Representative Directors: Jun Iinuma, Eric Whiteway)—a joint development partner in the field of OCR in The AOR™—with the objective of strengthening collaborative efforts with them to develop new services.

1. The AOR™
For hand–written documents and non–standard forms, which are difficult to automate as they require manual data–entry, this solution automates and streamlines overall operations through the highly accurate digitization of character information using AI, OCR, and RPA technology. Based on the results of proof–of–concept testing on requests for direct debit from accounts, manual data–entry was reduced by approximately 80%.
The AOR™ will also be introduced within Mizuho sometime in FY2019.

2. Proof–of–concept testing with regional financial institutions
With the prospect of utilizing The AOR™ at Mizuho, we have begun to introduce it in regional financial institutions. We are currently conducting proof–of–concept testing at six banks to verify the effectiveness of The AOR™. These banks include, The Gunma Bank, Ltd.; The Chiba Kogyo Bank Ltd.; The 77 Bank, Ltd.; The Daishi Bank, Ltd.; The Kagoshima Bank, Ltd.; and The Higo Bank, Ltd. The aim of the testing is to enhance cooperation with regional financial institutions and improve the productivity of financial institutions overall.

3. Collaboration with Primagest, Inc. and RKK Computer Service Co., Ltd.
We have strengthened our framework for providing The AOR™ to regional financial institutions through collaborative efforts with Primagest and RKK Computer Service which provide services related to back office workflow systems in financial institutions. By collaborating with these companies, which have a significant market share in systems used in financial institution operations related to direct debits from accounts, we will be able to facilitate the smooth introduction of The AOR™.

4. Cogent Labs Inc.
The strength of Cogent Labs lies in its development of services using cutting–edge practical AI applications. It provides services mainly to leading companies in Japan. These services include "Tegaki"—a service that digitally converts handwriting into data; "Kaidoku"—a natural language understanding engine that identifies key insights from text–based data; and "Time Series Forecasting (TSF)"—an application that generates forecasts for financial markets and other fields. We have worked together with Cogent Labs using "Tegaki" in the field of OCR in The AOR™ and will continue to enhance AI–based solutions by strengthening our collaborative efforts with them.

PDF for print (PDF/77KB)