T&D Financial Life Begins Verification for AI Operator Implementation with RightTouch's "QANT Speak". Inquiry-based Branching Accuracy Records 99.3%.

RightTouch Co., Ltd. (Head Office: Shinagawa-ku, Tokyo, Representatives: Shuhei Nomura / Daito Nagasaki, hereinafter "RightTouch") announces that T&D Financial Life Insurance Company (Head Office: Minato-ku, Tokyo, President: Kaya Morinaka, hereinafter "T&D Financial Life") has commenced a Proof of Concept (PoC) for the "QANT Speak" service in September 2025, targeting its telephone channel (My Number related dial). This verification is proceeding to assess accuracy in listening and inquiry branching, paving the way for AI operator implementation. T&D Financial Life is establishing a foundation for balancing customer experience and operational efficiency, anticipating future inquiry concentration. Furthermore, to enhance T&D Financial Life's customer experience (CX), "QANT Web" and "QANT VoC" are also being introduced alongside "QANT Speak" as announced in this press release. Related link: "T&D Financial Life Introduces RightTouch's "QANT Web" to Improve Official Website FAQ, Promoting Self-Resolution and Inquiry Reduction Through Proactive Answer Presentation" Related link: "T&D Financial Life Introduces RightTouch's "QANT VoC" to Streamline Complaint Extraction and Internal Report Creation Through VoC Analysis" Background In the life insurance industry, while digital channels are increasingly utilized, the telephone channel remains a crucial point of customer contact. Inquiries regarding contract details and procedures are particularly important, where customers can confirm directly with an operator, providing peace of mind. Accurate listening and appropriate inquiry branching significantly influence the quality of guidance provided. In recent years, with the rise in inquiries and staffing challenges, there has been growing interest in AI utilization within contact centers. During periods of concentrated inquiries, issues such as increased wait times and operational load become problematic, drawing attention to the potential of AI for initial responses. However, the life insurance industry requires careful verification from the perspectives of response accuracy and control. T&D Financial Life is facing a significant management challenge in preparing for future inquiry increases, with a large number of insurance policies set to mature simultaneously in 2028. In light of this situation, and with the goal of implementing AI operators, they are incrementally verifying the accuracy of AI-driven listening and inquiry branching within a limited target channel. The aim is to establish an operational model capable of maintaining stable responses even during future periods of high inquiry volume. Verification Content * **Verification of Listening and Inquiry Branching Accuracy** The company is evaluating whether inquiries are being appropriately routed to the correct department based on AI's listening results, after pre-organizing anticipated inquiry patterns. Actual call data is being used to verify if the classification is as expected. * **Verification of Hybrid Operations Combining AI and Humans** The company is verifying an operational model where AI handles initial listening and summarization, and humans then confirm and supplement these results. By designing the system to allow for appropriate human involvement rather than relying solely on AI, the operational feasibility is being confirmed while maintaining guidance quality. * **Verification of Response Scope Based on Predefined Rules** The system is being verified where AI executes tasks from listening to inquiry branching based on predefined inquiry scopes and routing rules. The design includes human intervention for unexpected cases or difficult-to-judge content to ensure operational control. Verification Progress Internal verification targeting specific call reasons demonstrated 99.3% accuracy in inquiry branching based on AI-IVR (Interactive Voice Response) listening results. Currently, the company is proceeding with expanding the verification scope and refining the operational design for field trials at customer touchpoints. Future Outlook Based on these verification results, T&D Financial Life aims to gradually expand the scope of AI operator utilization to establish a stable response system for the telephone channel. Moving forward, they will comprehensively utilize customer data accumulated from both telephone and web channels, reflecting inquiry trends and customer pain points identified through QANT VoC into knowledge base development and response design. The developed knowledge base will be used for self-service support via QANT Web and applied to telephone responses via QANT Speak, thereby achieving consistent guidance across channels. Through these initiatives, they will progressively advance towards realizing an AI contact center by appropriately designing AI for initial responses and human agents for specialized tasks. By creating a cycle encompassing analysis, knowledge development, contact point application, and effectiveness verification, they will support the construction of a customer response system capable of maintaining quality even during future periods of increased inquiry volume. RightTouch will support CX improvement by linking QANT Web, QANT VoC, and QANT Speak, providing integrated support from inquiry analysis and knowledge creation to web and telephone response application. Comments **Kunihiko Kaku, Executive Officer, Contract Service Department / Contact Center Department, T&D Financial Life Insurance Company** "We are approaching a period where a large number of insurance policies will mature in the next few years, making preparation for future inquiry increases a significant challenge. With this in mind, we have begun verification for the implementation of AI operators. Initially, we will incrementally verify the accuracy of AI-driven voice listening and inquiry branching, as well as operational effectiveness, in an environment with limited target channels. This will enable us to maintain stable response quality even during future periods of increased inquiries, while simultaneously advancing operator roles and establishing a new customer contact model that further enhances customer experience." **Shuhei Nomura, Representative Director, RightTouch Co., Ltd.** "With T&D Financial Life, we are jointly advancing initiatives to enhance customer understanding and contact point design by linking QANT Web, QANT VoC, and QANT Speak, aiming for sustained CX improvement. The telephone channel is a critical touchpoint for consultations regarding contract details and procedures, where accurate listening and appropriate routing are paramount for response quality. In this project, AI will handle the listening and summarization of inquiries, and the results will be confirmed and supplemented by humans, thereby achieving efficiency while maintaining quality. Furthermore, by utilizing response logs and continuously reviewing knowledge data that supports AI judgments and guidance, we aim to gradually expand the scope of AI's capabilities and establish a response model that can handle future increases in inquiries." ■ QANT's Value Proposition "QANT" supports the multifaceted implementation of AI across various customer support operations and customer touchpoints, while also connecting and optimizing the entire workflow. Labor-intensive tasks such as problem analysis, proposal creation, and knowledge development are automated by "AI," allowing human resources to concentrate on areas requiring human judgment, such as content verification, decision-making, and interpersonal communication. This reduces workload and enables greater focus on creating a better customer experience. Furthermore, by accumulating customer support data continuously through this cyclical process, an ideal PDCA cycle is generated that enhances AI accuracy. Regarding "QANT Speak" QANT Speak is a self-improving AI operator that becomes smarter with use and expands its scope of response. It is equipped with a mechanism to continuously enhance its accuracy and response range by accumulating and utilizing logs from AI operator and human agent interactions, thereby completing the loop of response → analysis → knowledge improvement. About RightTouch Co., Ltd. With the mission "to liberate all people from negative experiences and unlock their potential," RightTouch develops and provides "QANT (Quanta)," a platform for "Customer Support Automation" that enables a continuous loop of PDCA through human-AI collaboration. Through multiple products for VoC analysis, web support, and contact center operations, they achieve dramatic improvements in CX/EX alongside labor reduction, supporting the transformation of customer support (CS) for enterprise companies across various industries including finance, infrastructure, and retail. RightTouch is a startup spun off from Plas Co., Ltd. (Tokyo Stock Exchange Growth 4165). Name: RightTouch Co., Ltd. Location: Meguro Techno Building 2F, 4-31-18 Nishi-Gotanda, Shinagawa-ku, Tokyo Representative: Shuhei Nomura / Daito Nagasaki Established: October 27, 2021 Business Description: Development and provision of the customer support platform "QANT". Company URL: