Skip to main content
Roy Huang

Profile

MEng student in Electrical and Computer Engineering at the University of Waterloo with a BEng in Computer Science and Artificial Intelligence from HKUST. Skilled across agentic AI, large language models, and computer vision, with hands-on expertise in building robust evaluation benchmarks and deploying intelligent systems. Adept at bridging machine learning research with practical software engineering to deliver scalable, high-performance applications for real-world impact.

Education

University of Waterloo

MEng in Electrical and Computer Engineering (Co-op)
Sep 2026 to Present | Waterloo, Canada
  • Relevant coursework: Responsible AI, Advanced Image Processing

National Taiwan University

MSc in Computer Science and Information Engineering (Withdrawn)
Sep 2025 to Jan 2026 | Taipei, Taiwan
  • GPA: 3.96
  • Relevant coursework: Cognitive Computing, Reinforcement Learning

Hong Kong University of Science and Technology

BEng in Computer Science & Extended Major in Artificial Intelligence
Sep 2021 to Jul 2025 | Hong Kong
  • GPA: 3.74
  • Awards: First Class Honors, Dean's List x 4, Champion of Best CPEG Final Year Project.
  • Relevant coursework: Artificial Intelligence, Machine Learning, Computer Vision, Natural Language Processing, Data Structures, Software Engineering, Operating Systems, Computer Architecture, Parallel Programming, Cloud Computing

Technical University of Denmark

Exchange Student
Sep 2023 to Jan 2024 | Copenhagen, Denmark
  • Relevant coursework: Algorithms, Machine Learning and Data Mining, Image Processing, Data Security

Experience

Delta Electronics

Intern (Advanced Technology II)
Jul 2025 to Dec 2025 | Taiwan
  • Built the ETOM benchmark to evaluate agentic orchestrators over a 2,000+ tool pool, designing 5 difficulty levels: L1 single-turn (explicit server/tool), L2 single-turn (ambiguous with Equal Function Sets ground truth), L3 multi-turn intra-server planning, L4 multi-turn cross-server planning (with EFS), L5 "impossible" tasks for capability-gap detection.
  • Researched and benchmarked orchestrator architectures (ReAct, MCP-Zero, ToolShed) on ETOM, analyzing planning quality, tool selection, and multi-turn reasoning trade-offs under large tool sets.

Cathay Securities Investment Trust

Intern (AI Team)
Jul 2024 to Aug 2024 | Taiwan
  • Developed an evaluation framework for LLMs to assist in selecting optimal open-source LLMs for local deployment.
  • Utilized the DeepEval framework for comprehensive assessments, incorporating benchmarks from the G-Eval framework.

Descamps Technology

R&D Engineer Intern
Jun 2023 to Jul 2023 | Hong Kong
  • Designed schematic and 2-layer PCB for the single-sensor (photodiode) prototype enabling stable sampling.
  • Built calibration dataset with 258,048 samples. Achieved 96.77% accuracy with moving-average dynamic thresholds.
  • Engineered a zero-loss parallel data path on ESP32 using double buffers and semaphores, reducing processing latency.

Cathay Century Insurance

Intern (Customer Analytics Section, DDT)
Jul 2022 to Aug 2022 | Taiwan
  • Intern Project: Identify target customers for summer vacation insurance.
  • Used SQL to analyze past customer data and select target customers.
  • Analyzed customer purchase behavior to inform future product promotions.

Publications

Skills

  • Programming Languages: Python, C/C++, Java, SQL, MIPS, LaTeX, Bash, JavaScript
  • Frameworks & Libraries: React, PyTorch, TensorFlow
  • DevOps & Infrastructure: Git, Docker, Kubernetes, AWS
  • Parallel Computing: CUDA, OpenMP, Pthreads, MPI
  • Other/Soft Skills: Technical Presentations, Agile/Scrum, Bilingual (English, Mandarin)