Nextremer Co., Ltd.
Tokyu Corporation
Shibuya Tourism Association General Foundation
Nextremer Co., Ltd. (hereinafter, Nextremer) and Tokyu Corporation (hereinafter, Tokyu) have received the "Special Award for Sustainable Implementation" and the "Cooperating Organization Special Award HDI-Japan Award" (hereinafter, these awards) in the "AI Agent Development for Social Issue Resolution Utilizing Domestic Foundation Models, etc." domain, one of the three domains, of the prize-based program "GENIAC-PRIZE" (hereinafter, this program) hosted by the Ministry of Economy, Trade and Industry and NEDO (New Energy and Industrial Technology Development Organization).
This program utilizes the NEDO prize-based program (NEDO Challenge) system, with a total prize money of approximately 800 million yen, targeting social issues, government agencies, and safety across three domains where solutions are sought through generative AI services. It aims to promote development and verification by diverse entities in various regions, accelerating the utilization and social implementation of generative AI applications. Among the domains in this program, which was held for the first time this year, the "AI Agent Development for Social Issue Resolution Utilizing Domestic Foundation Models, etc." domain received a total of 114 applications (Manufacturing: 58, CS: 56). Eight applicants per theme were selected as final candidates, and the winners were decided through a final screening via pitch presentations held on March 24, 2026.
For this award, the verification (hereinafter, this verification) conducted at two tourist information centers, "SHIBU HACHI BOX" operated by Shibuya Tourism Association General Foundation and "WANDER COMPASS SHIBUYA" operated by Tokyu, utilizing the conversational AI service "AI MINARAI" provided by Nextremer, aiming to improve the efficiency of tourist guidance and standardize guidance skills, was highly evaluated, leading to these awards. The results of this verification are as follows:
<Features of this verification>
By having the AI handle routine questions and converting the time saved by the AI into "omotenashi" (hospitality), we achieved reduced counter congestion and improved customer experience quality (Response workload: 22.4% reduction).
Utilizing a domestic LLM (Large Language Model) as an evaluation engine, the answer accuracy rate improved from 76.0% to 94.5%.
Constructed a conversational simulator capable of realizing various dialogue cases through the collaboration of multiple agents, conducting thousands of tests with human-like natural conversation, thereby improving data quality and streamlining data creation.
An app-free AI agent that inbound tourists can use simply by scanning a QR code.
Award Ceremony
Image of Chat with Generative AI
By continuing to explore the potential of generative AI, we aim to provide new value that meets diverse needs and to deploy convenient and innovative conversational AI services...
FACT BOX
- Source: PR Times
- Category: News