Master of Science in Agentic AI Systems

Aims and Characteristics

Programme Aims
The Master of Science in Agentic AI Systems (MScAS) programme moves beyond the training of isolated AI algorithms to the engineering of complete, autonomous ecosystems.  It prepares students to bridge the gap between abstract AI models and real-world applications by integrating advanced reasoning with robust hardware and software infrastructure. Covering topics such as large models, embedded systems, robotics, and cloud architecture, the curriculum enables students to design end-to-end AI agents capable of perceiving, reasoning, and executing complex tasks. Graduates will be equipped to transform model intelligence into practical system intelligence for real-world impact.

The objectives of the programme are listed below:
 

• To equip students with the advanced knowledge and practical skills to design, implement, and orchestrate the end-to-end systems for autonomous agents. This includes integrating reasoning, decision-making, and interactive capabilities across both software and hardware infrastructures;
• To cultivate interdisciplinary leaders for academia and industry by providing deep expertise in the convergence of agentic AI, robotics, embedded systems, and intelligent automation. Graduates will be prepared to drive innovation in the development and application of these technologies responsibly and effectively; and
• To address a critical market gap by supplying Hong Kong with high-quality personnel capable of pioneering the next generation of intelligent, autonomous systems, thereby meeting the urgent and growing demand for expertise in agentic AI.

Programme Characteristics

• Full-Stack AI Engineering for Real-World Impact
  Our curriculum is designed to dissolve the boundaries between code and machinery, emphasising the seamless integration of hardware and software systems. Students master full-stack AI engineering, learning to optimise intelligent agents for both cloud environments and deployment on edge devices and robotic platforms.

• Zero-to-Physical Capstone: Bridging Software and Hardware

  Experience hands-on learning with our signature capstone project, where students develop agentic workflows that link software intelligence to physical hardware, translating sensor data into real-world actuator movements.

• Hybrid Learning Environment: Powerful Tools, Practical Experience
  Benefit from a unique blend of high-performance computing clusters and state-of-the-art electronics labs. Our hybrid infrastructure provides the resources and mentorship needed to architect unified systems—where sophisticated AI algorithms seamlessly drive mechanical action, preparing students for the future of intelligent automation.