李育豪

專長

Master's in Industrial Management from National Central University; Bachelor's in Computer Science and Industrial Engineering & Engineering Management from National Tsing Hua University, combining software engineering with industrial engineering training.

Manufacturing data integration and getting AI into production use: plain-language report queries, data de-identification and on-premise deployment, floor dashboards with closed-loop abnormality alerts, ERP / MES data interfacing, SMB real-time monitoring with PLC data acquisition, and n8n process automation.

Certified by Taiwan's Ministry of Economic Affairs iPAS programme as an AI Application Planner (Data Analytics), Intermediate (certificate A-J21-0177-2025) and an Information Security Engineer, Intermediate (certificate B-S21-1429-2026); also holds iPAS Foundation certificates in AI Application Planning and Smart Production, plus n8n Level 1 and Level 2. Completed ITRI's AI implementation planning programme and the Industrial Competitiveness Advisory Corps seed training (Southern Region, 6th cohort).

Toyota / 國瑞體系經驗

About seven years in information systems and project work at Kuozui Motors (the Toyota system), covering HR system implementation, supplier and quality control, and the introduction of machine vision and patrol-inspection mechanisms. Previously a section manager at Innolux, responsible for manufacturing process management and reporting system development.

代表案例

Named inventor on ROC Invention Patent I754131, "Process Management System and Process Management Method" (co-owned with G-SHANK Enterprise Co., LTD.); also holds utility model patents for an inventory management system, a kanban management system and a data transmission system, and an invention patent for an "Intelligent Massage System and Control Method" (US patent pending)

Principal investigator on an SBIR project for the Small and Medium Enterprise and Startup Administration, MOEA (intelligent natural-language query and dynamic code generation data management platform; completed).

First work out which part does not need a system at all. Then build only what is genuinely left.