2025-03-25
As Large Language Models (LLMs) become a pivotal driving force in digital transformation, supportive policies by the Taiwanese government have accelerated the deployment of innovative solutions. Utilizing Retrieval-Augmented Generation (RAG) technology, LLMs can instantly retrieve information from internal databases and external literature, significantly enhancing the accuracy and reliability of content generation. Furthermore, multimodal model integration enables AI to handle diverse data formats, including text, tables, and images, thereby expanding application potential.
Through cross-institutional collaboration, Industrial Technology Research Institute (ITRI) has successfully developed numerous advanced technologies, such as regulatory compliance reviews and intelligent document management systems. These systems automatically analyze document content, provide accurate compliance recommendations, and ensure that document format and content meet standardized requirements, significantly enhancing document generation quality and management efficiency. Generative AI (GAI) technologies have also been widely applied in medical education and drug discovery. For instance, in medical training, GAI can simulate standardized patients, creating highly realistic digital teaching platforms that enhance clinical training outcomes. These advancements streamline regulatory reviews and research processes, substantially reducing costs as well as expanding globally into diverse innovative applications, thus providing tangible benefits and enhancing industry competitiveness.
1. Automated Report Generation - Regulatory Compliance Review Management
As global financial regulations become increasingly stringent, Generative AI (GAI) has become essential for enhancing compliance reviews and risk management efficiency. Leveraging Retrieval-Augmented Generation (RAG) technology, GAI swiftly extracts critical information from internal databases and external regulatory documents, accurately interprets complex financial regulations, terms, and contract contents, and rapidly summarizes key points and performs risk assessments. Multimodal AI models further broaden the applicability by processing various financial documents, tables, and images simultaneously, generating precise recommendations compliant with regulatory requirements, thus markedly improving document quality and operational efficiency.
Based on this technology, Industrial Technology Research Institute’s (ITRI) Information and Communications Research Laboratories developed an end-to-end (E2E) intelligent contract management solution that integrates AI into audit workflows. Coupled with anomaly detection techniques, the system automatically identifies potential risks in financial statements and transaction records, greatly enhancing audit accuracy and operational efficiency. Additionally, the real-time regulatory monitoring system encompasses key functions such as regulatory data collection, clause analysis, compliance checking, risk identification and improvement recommendations, and regulatory update monitoring, enabling financial institutions and supply chain partners to swiftly adapt to regulatory changes and minimize compliance risks. This technology has successfully shortened contract drafting and review times, reducing legal resource requirements by over 50% and contract processing time by over 40%, and has been adopted by several major financial institutions.
The implementation of this technology significantly improves regulatory review efficiency, reduces compliance costs, and further strengthens audit accuracy and operational effectiveness, enabling financial institutions and large corporations to maintain strong competitiveness in global markets while driving digital and intelligent transformation of compliance management.
Caption: Compliance Review Management
2. Digital Medical Education - Standardized Patient Simulation
With the rapid advancement of digital medical education, digital dental education systems employ simulated clinical scenarios, intelligent Q&A modules, and learning outcome analysis to provide comprehensive and efficient training for dental students and healthcare professionals. The system integrates standardized patient (SP) interactive technology with voice-enabled Q&A functions, allowing students to practice consultation skills in highly realistic clinical scenarios, improving clinical responsiveness and diagnostic decision-making quality. The core of this system employs Large Language Models (LLMs) capable of intelligently matching students' consultation dialogues with standardized patient responses, combined with digital medical records, to provide personalized learning recommendations that strengthen clinical knowledge and diagnostic reasoning skills.
Furthermore, the platform developed by Industrial Technology Research Institute’s (ITRI) Information and Communications Research Laboratories can automatically generate detailed learning reports, highlighting gaps during consultations and analyzing diagnostic recommendations. It also supports highly realistic clinical simulations, including dental implant planning and 3D tooth arrangement analysis, significantly enhancing students' clinical operational abilities. Students can log into the system using personal accounts for examinations and simulation exercises, where AI provides real-time answer validation, performance assessments, and personalized improvement recommendations, effectively enhancing professional skills and clinical decision-making. Clinical testing and user interviews reveal system satisfaction ratings as high as 8 out of 10, and it has already been successfully deployed in three hospitals in Taiwan with support from ITRI’s Biomedical Technology and Device Research Laboratories. Additionally, discussions are underway with Stanford Medical School to further advance digital medical education globally and expand into broader medical training fields.
In the future, this technology will enhance dental and medical education while simultaneously being adaptable to professional communication training across various industries—including customer service, demand exploration, and business communication. It offers efficient and intelligent learning solutions, further accelerating the integration of digital education with AI technologies.
Caption: Illustration of Digital Medical Education
The application of Generative AI technologies has successfully expanded across various sectors, demonstrating strong capabilities in data processing and innovation from regulatory compliance management to medical education. AI adoption enables enterprises to streamline processes, reduce costs, and enhance market competitiveness. With continuous technological advancements, generative AI is expected to expand, bringing greater added value to multiple industries and accelerating global digital and intelligent transformation.
Source： Dr. Hsiang-Wei Hu, Information and Communications Research Laboratories, Industrial Technology Research Institute (ITRI)
About Us
InvesTaiwan InvesTaiwan Service Center Department of Investment Promotion, MOEA Department of Investment Review, MOEA Contact TaiwanPlanning
Reasons to Invest Overview Key Industries for Investment Promotion Incentives Important Policies Success Stories
