Hey! I'm Yixin ("Isshin"), a computer engineer based in Toronto, Canada. I build human-centred AI systems at the intersection of software engineering, product design, and AI research. My work focuses on human-AI interaction, multi-agent systems, learning science, semantic technologies, and AIOps.
I worked as a Data Scientist at Royal Bank of Canada (RBC), where I focused on developing agentic AI assistants for the commercial credit space leveraging DSPy, LangGraph, and tool-augmented LLMs. Over the summer of 2025, I served as one of the team leads for RBC Amplify, mentoring a team of four students to rapidly prototype an AI assistant that uncovers actionable commercial client insights using LangGraph, RAG, unsupervised anomaly detection, SHAP, and Text-to-SQL. My previous experience spans building event-driven data pipelines and co-developing a patent-pending agriculture carbon emission model into production at RBC.
Outside of work, I was fortunate to conduct two years of part-time research at USC's Information Sciences Institute advised by Prof. Craig Knoblock, focusing on agentic structured data extraction from long documents using LLM agents and knowledge graph construction. Previously, I worked with Prof. Shurui Zhou at the University of Toronto, where I contributed to developing a Jupyter Notebook extension that visualizes data science workflows to make knowledge sharing more efficient among practitioners.
I'm interested in how human-AI interaction patterns impact individual learning, including epistemic agency, metacognitive thinking, and procedural skill formation. I'm also interested in how teams of humans and AI agents collectively create and share knowledge at scale, and what conditions enable outcomes neither could achieve independently. When I'm not building or researching, I enjoy drumming, running, writing, reading, and spending time in nature.