GMO Aozora Net Bank and University of Tokyo's Morikawa-Narisue Laboratory Launch Full-Scale Joint Research
GMO Aozora Net Bank and the University of Tokyo Graduate School of Engineering Morikawa-Narisue Laboratory have officially launched joint research towards the construction of innovative credit risk assessment models utilizing AI and machine learning.
In other news, TRABOX has launched a new service called "Finto Invoice Payment", which significantly reduces the burden of accounting operations for corporate clients by using "BaaS by GMO Aozora" provided by GMO Aozora Net Bank to complete the entire process from invoice receipt to payment in a one-stop solution.
Joint Research on Innovative Credit Risk Assessment Model Construction Using AI and Machine Learning
This research aims to discover evaluation elements that could not be captured by conventional credit risk assessment models by combining the advanced knowledge and technology in the machine learning field possessed by the University of Tokyo Morikawa-Narisue Laboratory with the financial data accumulated by GMO Aozora Net Bank, utilizing AI and the tacit knowledge of financial institutions.
While initial investigations have been conducted since May 1, 2025, this announcement marks the transition to the research and development phase of new credit risk models based on customer clustering that does not depend on financial data, entering the full-scale research stage of the project. Focusing on the potential of diverse customer attribute data held by GMO Aozora Net Bank, the partners will challenge the visualization and quantification of credit risk that is not bound by conventional frameworks.
Background for Starting Joint Research
In recent years, advances in data-driven technology have promoted the digitization of trading information and consumer behavior data, expanding the possibilities for creating new value using such data. Accordingly, the strategic and effective utilization of held data has become an urgent issue for financial institutions to respond to rapid changes in the economic environment and diversifying customer needs.
Furthermore, with the advancement of FinTech, the utilization of non-financial data in addition to financial data is accelerating, and analysis beyond conventional frameworks is becoming possible in the field of credit risk assessment. Additionally, various methods including deep learning have been put into practical use in machine learning models, further improving analysis accuracy and flexibility.
Through the advanced fusion of such technology and data, the construction of new credit risk assessment models that can accurately grasp risk factors that could not be captured by conventional methods is expected at financial institutions and other organizations.
GMO Aozora Net Bank, which advocates being the "No.1 Technology Bank," has partnered with the University of Tokyo Morikawa-Narisue Laboratory, which has outstanding research achievements in the AI and machine learning fields, to work on constructing more advanced and sophisticated credit risk assessment models using the laboratory's cutting-edge technology and knowledge, as well as the abundant financial data accumulated by GMO Aozora Net Bank.
Through this joint research, GMO Aozora Net Bank aims to establish credit risk assessment models that can flexibly respond to rapidly changing economic environments and provide optimal financial services that achieve advanced risk management without compromising customer convenience.
TRABOX Launches "Finto Invoice Payment" Service - One-Stop Processing from Invoice Receipt to Bank Transfer
TRABOX has launched a new service called "Finto Invoice Payment", which significantly reduces the burden of accounting operations for corporate clients by using "BaaS by GMO Aozora" provided by GMO Aozora Net Bank to complete the entire process from invoice receipt to payment in a one-stop solution.
In addition to reducing the workload of tedious invoice receipt and complex data entry, it also contributes to reducing transfer fees depending on the financial institution currently in use.
Background for Launching "Finto Invoice Payment"
"Finto," which underwent a brand renewal in February 2025, is a financial service consisting of "Finto Invoice Receipt" that enables one-stop management from invoice registration to settlement, and "Finto Card Deferred Payment". As of July 28, 2025, "Finto Invoice Payment" will be newly added to the service lineup.
Previously, TRABOX has provided "Finto Invoice Receipt" to support business efficiency through invoice management and automatic creation of journal entries and transfer data. However, in actual bank transfer processing, the hassle of importing created transfer data into internet banking still remained.
Furthermore, according to a survey of TRABOX members regarding bank transfer fees, 52.5% responded that they incur fees of JPY 300 or more per transaction, and 24.5% incur fees of JPY 500 or more per transaction, creating a significant cost burden for companies.
Additionally, challenges have been raised regarding the complexity of management due to inconsistent timing and formats of invoice arrivals, requiring significant time for manual data entry and content verification.
In response to this situation, TRABOX is launching "Finto Invoice Payment" as a new service to reduce the hassle and costs of bank transfer operations for corporate customers and achieve further efficiency and automation of overall invoicing operations.
What is "Finto Invoice Payment"?
"Finto Invoice Payment" automatically collects invoices issued and sent by business partners (suppliers, subcontractors, etc.) to Finto users (implementing companies) by sending them to a dedicated email address. AI reads and digitizes the data with high accuracy for speedy processing. The digitized invoices are automatically linked to transfer processing, enabling end-to-end handling from invoice receipt to bank transfer.
By completing the entire series of operations from invoice receipt to data entry, verification, and transfer processing in a one-stop solution—previously done manually—it reduces the risk of input errors and forgotten transfers while significantly reducing the workload for accounting staff.
Three Implementation Benefits of "Finto Invoice Payment"
1. Complete everything from invoice receipt to transfer with Finto alone!
Simply receive invoices from partner companies via a dedicated email address, and AI automatically reads invoice content and executes transfer processing automatically.
2. Reduce transfer costs with a flat rate of ¥240 (tax included) to any bank!
The cost for transfers through "Finto Invoice Payment" (including transfer fees + service usage fees) is ¥240 per transaction from invoice receipt to transfer, regardless of the destination bank, reducing both the hassle and cost of payment operations.
3. Freedom from tedious work and mistakes like input oversights and forgotten transfers—significant workload reduction!
Eliminates the hassle of tedious invoice data entry, logging into multiple services, and data migration, while preventing mistakes like input oversights and forgotten transfers, significantly reducing payment processing workload. This allows accounting staff to focus on core operations.
Norbert Gehrke
