Developed a course recommendation system that leverages large language models and retrieval-augmented generation to transform academic advising at large universities. The system processes natural language queries about academic interests and generates personalized course recommendations across disciplines, complete with detailed rationales and confidence assessments. By bridging the gap between how students express interests and formal course descriptions, it addresses key challenges in course discovery, particularly for new students navigating thousands of options. The project demonstrates technical expertise in LLMs, embedding-based retrieval, and bias testing while solving a real institutional need. Currently being piloted through the University of Michigan's Atlas platform to enhance academic planning and equity in course access.
This was a research replication project I worked on for a class. We replicated the paper "Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methods". I primarily led understanding and presenting the mathematical ideas behind post-hoc explainer methods like LIME and SHAP and proposed our extension idea of combining the two methods to make a more resistant method. I developed and evaluated our proposed method.
A sensitivity analysis on ABBA, a netlogo agent-based model developed by Jorge A. Chan-Lau that models the banking system. I investigated the sensitivity of bank credit and liquidity failure with respect to saver's withdrawl rates and recovered the common sense idea that higher withdrawl rates (lower liquidity) leads to greater liquidity failures. This basic fact helps validate the modeling strengths of ABBA.