Center for Artificial Intelligence and Data Science
Research Field
Yen-Jen Chen
Associate Professor & Director of AI Center
Over the past five years, Professor Yen-Jen Chen has demonstrated profound insights in the research of AI Power Cloud, AIoT, and Large Language Model (LLM) technologies. He has accumulated rich practical experience in the industry, publishing 16 papers in international journals and conferences, successfully securing 10 invention patents, and leading 37 industry-academia collaboration research projects as the principal investigator or co-principal investigator. The total funding for these projects amounts to NT$28,893,000.
Since 2023, Professor Chen has been collaborating with Formosa Plastics Network (FPN) on a project titled "Construction of Computing Environment Deployment and Retrieval-Augmented Generation Model." In this project, he leads the development of a Retrieval-Augmented Generation (RAG) architecture based on the open-source Taiwan LLM. By leveraging Parameter-Efficient Fine-Tuning (PEFT) and Low-Rank Adaptation (LoRA) techniques, the project has enhanced the model's understanding of domain-specific knowledge, boosting the retrieval hit rate to over 95%. Professor Chen's research achievements in this area not only introduce generative AI to Formosa Plastics Group, providing a key solution that balances security and privacy, but also effectively improves the efficiency of reviewing corporate regulations and enhances model comprehension. This work lays a solid foundation for the future development of RAG technology.
History & Purpose
Ming Chi University of Technology established the AI center in August 2019, providing more than 339.9 square meters of space to set up central offices and classrooms, and investing more than 39 million dollars in computing equipment. The center is mainly dedicated to the research and development of Industrial Artificial Intelligence. The focus of research and development is to combine industry expertise (X) and artificial intelligence (AI) technology to solve industrial practical problems in the form of industry-university cooperation (Co-Op), and cultivate AI+X talents for the industry. Details are described below:
Industrial AI Technology
- Component Failure Prediction and Establishment of Preventive Maintenance Mechanism
- Material combination, scheduling, and manufacturing process optimization
- Design and integration of private cloud, sensors and IoT
- System architecture for industrial image recognition and quality inspection
- Exploration of business intelligence and management strengthening mechanism
- Industrial Safety Inspection and Accident Prevention
Talent Education
- School: Offer AI general courses and professional credit or degree courses
- Industry: Knowledge-oriented and project-oriented education/training courses
Large Language Model (LLM)
(1) Large Language Model Enhanced Text-to-SQL Generation
(2) Vision-Language-Action Model
(3) Retrieval Augmented Generation
[1] Wen-Chin Tsai, Yen-Jen Chen, "A method for detecting residual objects and an integrated system utilizing the method for residual object detection," 2025 Taiwan Innotech Expo Invention Competition, Silver Medal, Taipei World Trade Center Exhibition Hall 1, October 16-18, 2025.
[2] Yen-Jen Chen, Wen-Bin Chen, "Patrol Personnel Behavior Detection System and Method," 2024 Taiwan Innotech Expo Invention Competition, Silver Medal, Taipei World Trade Center Exhibition Hall 1, October 17-19, 2024.
[3] Yen-Jen Chen, Yu-Hsiu Yeh, Chun-Yuan Huang, Jen-Fu Yang, Wen-Bin Chen, "Instrument Value Calculation Method," 2023 Taiwan Innotech Expo Invention Competition, Gold Medal, Taipei World Trade Center Exhibition Hall 1, October 12-14, 2023.
Ph.D., Computer Science and Information Engineering, National Chiao Tung University.
Job Description
We are seeking a highly motivated AI Systems Engineer to design, develop, and deploy AI-driven systems. The ideal candidate should have strong experience in artificial intelligence technologies and system integration. Experience in Reinforcement Learning (RL) is a strong plus.
- AI Infrastructure: Design, implement, and maintain scalable GPU-accelerated computing clusters and storage systems.
- Model Deployment: Build robust inference engines and APIs to serve deep learning models with low latency and high throughput.
- Performance Optimization: Optimize model performance and resource utilization through techniques such as quantization, hardware-specific acceleration (TensorRT, CUDA), and efficient memory management.
- MLOps & Automation: Develop end-to-end machine learning pipelines (CI/CD) for automated training, evaluation, and seamless deployment.
- System Intelligence (The "Plus"): Leverage Reinforcement Learning (RL) to optimize system-level decision-making, such as dynamic resource scheduling, autonomous control, or hyperparameter tuning.
- Collaboration: Partner with AI researchers to transition experimental algorithms into reliable services.
Preferred Intern Educational Level
PhD student in Computer Science, Artificial Intelligence, Data Science, or a related field. Please note that we are unable to accept applications from undergraduate or Master's students at this time.
Skill sets or Qualities
- Programming: Strong proficiency in Python Rust for system-level development.
- Frameworks: Deep experience with deep learning libraries such as PyTorch.
- Containerization & Orchestration: Knowledge of Docker in Linux environments.
- Distributed Systems: Familiarity with distributed computing and training frameworks.
Preferred Qualifications:
- Simulation Tools: Experience with NVIDIA Isaac Sim, Unity ML-Agents, or specialized industrial simulators.
- Deployment: Knowledge of deploying RL policies on edge devices or real-time control systems.
- Publications: Research contributions to conferences.
Job Description
This role involves bridging AI research and production engineering. You will design, build, and optimize high-performance infrastructure to scale AI models across our platforms. Expertise or interest in LLMs is a strong plus.
- Inference Optimization: Optimize AI model inference (latency/throughput) using techniques such as quantization, pruning, and hardware acceleration (CUDA/TensorRT).
- Infrastructure Management: Design and maintain scalable GPU clusters and containerized environments using Kubernetes and Docker.
- MLOps Pipelines: Develop and automate end-to-end ML pipelines, including automated testing, continuous deployment, and real-time monitoring.
- System Integration: Collaborate with AI researchers to transform experimental models into reliable, production-ready APIs and services.
- Resource Allocation: Monitor and optimize GPU/CPU utilization to ensure cost-effective scaling of AI workloads.
Preferred Intern Educational Level
PhD student in Computer Science, Artificial Intelligence, Data Science, or a related field. Please note that we are unable to accept applications from undergraduate or Master's students at this time.
Skill sets or Qualities
Programming: Proficiency in Python; strong understanding of data structures and algorithms.
Deep Learning Frameworks: Solid experience with PyTorch or TensorFlow.
Cloud & DevOps: Experience with Docker.
System Design: Knowledge of distributed systems, RESTful/gRPC APIs, and database management.