Generative AI is reshaping the global industrial landscape at an astonishing pace. From finance and education to healthcare and manufacturing, countries around the world are accelerating the deployment of localized AI strategies to safeguard data sovereignty and strengthen innovation competitiveness. In response to this wave of challenges, the National Science and Technology Council (NSTC) is fulfilling the vision of “Sovereign AI” through the development of self-developed AI technologies. Today on November 11, the National Center for High-performance Computing (NCHC) of the National Institutes of Applied Research (NIAR) held the “TAIWAN AI RAP and TAIDE Highlights Presentation,” themed “Sovereign AI Driving Innovation, Diverse Industries Shaping a Smarter Taiwan.” The event showcased the latest achievements in localized generative AI development and real-world application cases.
The “TAIWAN AI RAP and TAIDE Highlights Presentation” drew enthusiastic participation from the industry, government, and academia. NSTC Minister and NIAR Chairperson Cheng-Wen Wu stated in his remarks that the government is promoting the “New Ten Major AI Projects.” The projects, built on digital infrastructure, key technologies, and smart applications, aim to accelerate Taiwan’s transition into a smart nation. Digital infrastructure emphasizes computing power, data governance, and sovereign AI, advancing regional balance, software-hardware integration, and talent cultivation. Smart applications focus on expanding Taiwan’s AI software industry, facilitating widespread adoption of AI tools across all sectors, and creating a smart-living ecosystem for every citizen so that AI becomes truly incorporated into the industry and daily life. Wu emphasized, “Within this national-level initiative, NCHC plays a pivotal role as the core provider of computing power, the R&D platform for localized AI, and a driver of interdisciplinary applications, underpinning Taiwan’s industrial upgrading and social innovation.”
Interdisciplinary Achievements Shows the Power of Local AI Innovation
The “TAIWAN AI RAP and TAIDE Highlights Presentation” brought together a wide range of showcase applications, fully demonstrating the deep integration between Taiwan’s AI technologies and local needs. From healthcare, education, agriculture and animal husbandry to enterprise operations and everyday life scenarios, the event highlighted how the TAIWAN AI RAP platform and the TAIDE language model had been adequately implemented across diverse sectors, exhibiting their strong innovation potential.
In the field of healthcare applications, Hsin-Cheng Liu, CEO of KETHY Inc., integrated the TAIDE model to launch the “診所通” smart search system, an app which accurately matches clinic services based on natural language queries, thereby enhancing the overall patient experience. Moreover, Po-Hsiang Lin, Doctor of Kaohsiung Veterans General Hospital, utilized the RAP platform to develop the “ER-Pulse Smart Shift Change System (ER-Pulse智慧交班平台),” achieving full automation of shift-changing procedures. The system shortens tasks that previously required several minutes of manual work to just a few seconds, marking a successful example of applying generative AI in clinical practice.
In the field of education, Tzu-Chiang Yen, CEO of Heartbot AI Inc., has developed an AI assistant platform powered by the TAIDE model to support secure deployment of smart-teaching tools, emphasizing contextual understanding and the incorporation of local impact. In the field of agricultural technology, Yu-Ting Hung, Division Chief at the Agricultural Technology Research Institute (ATRI), created an AI system called “Pig Matter Master (豬事問AI),” enabling farmers to obtain pig-health information and environmental data through simple text or voice input. The system significantly improves query efficiency and operational convenience, showcasing the tangible results of smart livestock management.
In the corporate application domain, Chih-Che Lin, General Manager of ShareGuru, showcased the “ShareQA AI Customer Service Platform for Knowledge Management (ShareQA AI知管客服平台).” The platform, the brainchild of domestically developed agentic AI technologies, integrates the TAIDE and other SLM models to provide SMEs with low-barrier SLM-based AI solutions. It strengthens enterprise knowledge management and service efficiency, accelerating smart transformation and serving as a model case for industrial upgrading.
In the area of technological R&D, the AI System Benchmarking and Tuning Lab at National Yang Ming Chiao Tung University (NYCU) introduced the “myPDA On-Premises GenAI System (myPDA全地端GenAI系統),” which creates AI avatars capable of understanding based on context. Its core technological device, the “Contextual Transformer,” accurately interprets idiomatic expressions and semantic nuances in traditional Chinese, enhancing response quality and advancing Taiwan’s linguistic sovereignty. Meanwhile, Twinkle AI and utuNote jointly developed two applications using the same platform. The former built Taiwan’s first ultra-small reasoning language model, while the latter created “AIYO (愛悠小幫手),” a user-friendly application designed for everyday life scenarios.
In addition, NCHC, in collaboration with the National Science and Technology Center for Disaster Reduction (NCDR), showcased the upgraded “Raining Bell App (落雨小幫手),” which boasts the latest AI-powered “Firefly Weather Q&A.” NIAR’s Science and Technology Policy Research and Information Center (STPI) also presented “iVoice,” a personalized voice-smart agent capable of rapidly converting speech to text, performing AI-driven analytical tasks, and engaging in interactive dialogue. These achievements not only demonstrate the deep integration of generative AI across diverse application scenarios, but also embody Taiwan’s advances in AI sovereignty, industry co-creation, and smart applications, marking an important step forward in Taiwan’s transformation into an AI-driven innovation island.
Linking Industry and Academia to Co-create and Expand the Ecosystem of Generative AI
The highlights presentation featured several distinguished speakers who provided critical industry insights and technological perspectives. Meng-Hsuan Chiang, Manager of Taiwan FinTech Alliance (TFA), kicked off with a keynote titled “A New Era of FinTech: An AI-driven Financial Ecosystem,” discussing how generative AI will reshape digital finance and demonstrating its potential to propel financial innovation. As a key partner in the development of the TAIDE model, Hen-Hsen Huang, Associate Research Fellow at the Institute of Information Science, Academia Sinica, delivered a specialized talk on “The Path to Technological Autonomy in Language Models.” He explored the evolving trends in LLM technologies and presented the training outcomes and international benchmarking of TAIDE’s localized models. A series of thematic talks and booth exhibitions were also arranged at the venue, showcasing real-world applications of generative AI across diverse industries.
NIAR Vice President Chun-Liang Lin noted that the institute is committed to building an R&D service platform that bridges the industry and academia by integrating its internal research capabilities. This approach aims to accelerate the transition of AI technologies from fundamental research to real-world industrial deployment. NCHC will continue to strengthen the integration of generative AI with high-performance computing (HPC) and expanding an application ecosystem centered on the TAIDE model. The goal is to bring AI-driven innovations into key domains such as education, healthcare, and manufacturing, propelling Taiwan towards a new developmental stage of smart and sustainable technologies.
NCHC Director General Chau-Lyan Chang stated that breakthroughs in generative AI have prompted countries across the globe to reassess the importance of AI sovereignty and technological autonomy. This presentation highlighted the latest achievements of two major platforms: the Taiwan AI RAP, a R&D platform for high-performance AI application, and the TAIDE, a reliable conversational AI engine. They demonstrate Taiwan’s steady progress in autonomous computing, AI platform technologies, and ecosystem integration. NCHC will continue to enhance the functional capabilities of the TAIWAN AI RAP platform and expand the application potential of the TAIDE model while encouraging autonomous integration of AI computing power and platforms so that industries can embrace AI innovations in a secure, sensible, and sustainable manner.
Source: NATIONAL INSTITUTES OF APPLIED RESEARCH