Financial Institutions

Reliance on the abilities of individual workers means that the quality of responses at bank call centers may vary depending on the skills and experience of the operator who takes the call. While increasing use has been made over recent years of generative artificial intelligence (AI) using retrieval-augmented generation (RAG)*1 to provide automated responses, this can suffer from hallucination problems when the user does not provide sufficient information.

A Hitachi technique for querying users overcomes the inadequate information problem by asking the user to provide the missing details. This allows the generative AI to better identify the intent of the question and to fill in missing information. The technique improves response accuracy. In verification testing with a bank*2, the technique improved the percentage of correct responses by 28 points from 58% to 86%. It is anticipated that this will reduce operator workloads and improve operational efficiency while also improving user satisfaction by shortening wait times.

The goal for the future is to generate new value by deploying the technique in areas beyond call center operations, such as supporting sales and review work and making system development more efficient.

1. A technique for providing more accurate responses by augmenting large language models with externally retrieved information
2. A technology trial conducted with The Keiyo Bank,Ltd. from May to September 2025

In our increasingly digital world, highly convenient biometric authentication that can provide reliable proof of identity is steadily becoming more widely adopted to ensure safe, secure, and convenient social infrastructure. Meanwhile rigorous management of personal information is needed to cope with threats that are becoming more sophisticated every year.

Hitachi’s Biometrics Integrated Infrastructure Service is based on its patented technology for public biometric infrastructure (PBI). By combining a diverse range of authentication methods with a high level of security that does not require the retention of actual biometric information, the service enables the use of biometric authentication in a wide range of applications.

During FY2025, Hitachi has extended the service to encompass the following industries and new areas of economic activity. These are in addition to existing applications in payment systems and login.

(1) Use at railway ticket gates and hotel check-in: The service has been deployed in SAKULaLa^*1, a joint platform by Hitachi and Tobu Railway Co., Ltd. that utilizes a biometric authentication system.
(2) Digitalization of transport concession tickets for the elderly: A demonstration project was run in partnership with Ishii, a town in Tokushima Prefecture.
(3) Issuing digital tourist tickets in combination with a beacon system: The service was used in tandem with JTB’s digital ticketing system in a demonstration project on Shodo Island in Kagawa Prefecture.
(4) Event entry management in tandem with Web 3.0 technology: The service was used to verify entry to a fireworks display in tandem with non-fungible tokens (NFTs)^* from Plat’Home Co., Ltd.

In the future, Hitachi intends to help bring about a society that enriches people’s lives by using the Biometrics Integrated Infrastructure Service as a bridge between the real and digital realms.

In April 2025, Hitachi launched a cloud-based low-code tool service that helps financial institutions accelerate the digital transformation (DX) of their client interactions. Offered in software-as-a-service (SaaS) form, the service utilizes Hitachi’s secure Internet banking platform^*1 running on Amazon Web Services*. It significantly reduces the amount of work that financial institutions need to devote to system configuration.

The service eliminates the need for multiple systems by providing a single platform for handling both in-person and remote customer interactions, enabling customers to undertake transactions wherever they are. It also incorporates proprietary low-code development and continuous integration/continuous delivery (CI/CD)^*2 functions from Hitachi that give service customers the ability to add or modify systems themselves. This speeds up the introduction of business process improvements.

The service also provides the infrastructure for institutions that use the service to develop their own practices. In the future, Hitachi intends to give a powerful boost to DX at financial institutions by further expanding the scope of service delivery.

The growing use of blockchain and the expansion of markets for digital assets such as crypto-assets or stablecoins are making financial transactions more global and much more immediate. Meanwhile, as scams and fraudulent transfer attempts become more ingenious, factors such as cost and workforce shortages place a limit on how far institutions can go with their existing in-house anti-money-laundering (AML) measures for digital assets. Especially in areas where the relationships between on-chain information and (off-chain) customer details are complex; this results in different institutions duplicating each other’s due diligence efforts and delays in responding to incidents that risk exacerbating damages.

In response to this challenge, Hitachi is establishing a joint AML center for the industry. This center will be available for use by crypto-asset exchanges, and financial institutions that issue or deal in stablecoins. The center will collate information on suspicious addresses and risk assessment logic that can be shared across the industry and use data sharing of information held by individual institutions to implement sophisticated monitoring and efficient due diligence processes. In anticipation of the coming era in which the trading of digital assets is an integral part of the social infrastructure, Hitachi is running a trial with participating companies to assess the efficacy and practicality of these measures with the goal of building a joint industry platform that can make AML more efficient and effective.

Hitachi’s know-how visualization and training solution addresses the problem of domain knowledge loss, an issue that is being exacerbated by the retirement of experienced staff and greater labor mobility. Based on the management of tacit knowledge and the SECI model^*1 for transforming it into explicit knowledge, the solution works by becoming part of the routine skills transfer process.

The first step is to extract tacit knowledge from existing documents and work logs and then formalize the knowledge held by individuals by categorizing it into skills, practices, processes, and human connections based on the Herrmann model^*2 of how the human brain works. The next step is to identify who will be involved in transferring each instance of the identified expertise and to establish a workflow for learning and updating. Once this is completed, interactive support by generative AI and auto-generated proficiency tests are used to incorporate the expertise into actual business operations. The solution can also visualize the expertise portfolio and its updating, providing an environment in which organizations can manage skills transfer in a systematic manner.

The solution operates on SharePoint* and helps maintain knowledge assets and improve business continuity by providing comprehensive support for everything from document collation to the transfer and adoption of workplace skills.