AI SOFTWARE · ML ENGINEERING · TECHNICAL PRODUCT · CANADA
From few-shot models on smartwatches to real-time family platforms and multimodal AI agents — I turn research-grade ideas into working, field-tested products.
This recruiter concierge grounds answers in verified portfolio evidence and maps a job description to the projects that support—or do not yet support—its requirements.
Not a chat wrapper: a bounded agent orchestrates function calls, hybrid semantic retrieval, verified evidence, job-requirement mapping, and grounded synthesis—with explicit failure and cost controls.
Portfolio Q&A, comparison, or job-fit assessment.
The agent invokes a strict, read-only evidence-search function.
Answer only from verified chunks and expose clickable sources.
SELECTED SYSTEMS / 01—05
Six systems. Different constraints. One consistent pattern: build the complete pipeline and test it with evidence.
01
Personalized behavior intervention, running on the wrist.
Built the end-to-end loop: accelerometer sensing, ten-example model customization, real-time inference, and just-in-time prompts on a smartwatch.
Journal paper · ACM Transactions on Computing for Healthcare · 2026
View case study →
02
A real-time coordination surface, designed inside family life.
Led the project end to end: translated situated family needs into a responsive multi-layer system, built its real-time client–server architecture, and evaluated how three blended families appropriated it in everyday life.
Master’s thesis · ACM CHI EA · 2026
View case study →
03
Orchestrated language, voice, video, personas, sessions, and token-aware memory in a full-stack conversational prototype.
CHI Best Paper HM (Top 5%) · China Daily—HK Edition
View case study →
04
A retrieval-augmented storytelling pipeline that grounds generated questions and answers in curated child-safety knowledge.
East China Normal University · Bachelor Thesis · 2023
View case study →
05
Built human–AI decision interfaces across two linked research programs and six controlled studies with 795 participants.
Hong Kong University of Science and Technology · CHI 2023 & 2024
View case study →THE ENGINEER BEHIND THE SYSTEMS
I work where machine learning, full-stack systems, and human behavior meet. My research background trained me to ask sharper questions; my engineering work taught me to answer them with systems people can actually use.
Software Developer (AI Systems & Backend), ContractMotiv Creative Labs ↗
Research Assistant (Web)Simon Fraser University ↗
Research Intern (Machine Learning & Wearable Systems)University of Washington ↗
Research Intern (Machine Learning, Multimodal AI & Web)Hong Kong University of Science and Technology ↗
Research Intern (Machine Learning, NLP & Web)Institute for AI Education, ECNU ↗
EDUCATION
B.Eng. in Computer Science & Technology
Highest Honors · Rank 1 / 117 · Shanghai Outstanding Graduate ↗CHOOSE YOUR COFFEE BREAK
Test your reflexes, or investigate some deeply suspicious office messages.
Hit as many signals as you can.
No résumés were harmed in the making of this game.
Who wrote it? Swipe ← Human · AI →
HUMANAITrust your office instincts.
LET'S BUILD SOMETHING USEFUL