Xinxin Yu

Master’s Student · AI & Data Science · University of Waterloo

I am a Master’s student in Computer Science at the University of Waterloo, specializing in data analysis, generative AI, and human–AI interaction. My work focuses on building end-to-end, data-driven systems that integrate large language models, automated evaluation, and interactive visualization to support real-world data science workflows.

I enjoy working at the intersection of analytical rigor and system design— connecting backend databases, AI model pipelines, and front-end interfaces to create explainable, efficient, and human-centered AI tools.

Research & Systems

My primary research explores how programmers can better evaluate, compare, and understand code generated by large language models in data science contexts.

DSCode Comparator (CGM)
Waterloo Insight Lab · Advisor: Dr. Anamaria Crisan

An interactive, production-level web system that enables users to evaluate, compare, refine, and diagnose LLM-generated data science code across multiple functional dimensions such as correctness, efficiency, and usability. The corresponding research paper has been accepted to ACM IUI 2026.

Cognitive Health Diagnostics AI
S AI Lab · Advisor: Dr. Frank Rudzicz

Developed hybrid LLM and machine learning pipelines for early dementia detection using linguistic, numerical, and speech-derived features from the Talk2Me (ISLAND cohort) dataset. Built an interactive Streamlit application to support interpretable predictions and exploratory analysis.

Publications

Technical Skills

Links