• Code and hardware design generation using large language models (LLMs), specifically in Verilog code, debugging, and hardware design, funded by Engineering Learning Initiative from Cornell University• Generated a data set consisting of waveforms and high‑level descriptions of Verilog code for inference context and corresponding Verilog modules using Python scripts, iVerilog, and OpenAI API, placing 9th in the NVIDIA LLM4HWDesign Contest.

• Developed and integrated spline‑based navigation with cubic Bezier curves in C++ for an open‑source autopilot software suite, optimizing mission flight time, improving energy efficiency, and enhancing mission planning accuracy and obstacle avoidance.• Engineered and implemented a robust database schema and reverse proxying configuration using PostgresSQL, Prisma, and NGinx, PM2 to develop a comprehensive full‑stack web application for alumni‑student‑sponsor relations.

A prestigious Jane Street Women in STEM Education (WISE) program.

Cynthia Shao and Runze Li, TPCselect: An R Package for Threshold Partial Correlation for Variable Selection. Developed an open‑source software package for variable selection via partial correlation that provides an alternative useful tool to regularization methods. Has 4,226 downloads since launch, and about 160 downloads per month.

Cynthia Shao and Runze Li, “TPCselect: An R Package for Threshold Partial Correlation for Variable Selection.”

Cynthia Shao (2022), “College Tuition, Gas Price, Inflation, and Stock Return”, presented in INFORMS (Institute for Operations Research and the Management Sciences) Annual Meeting, Indianapolis, IN.Applied various machine learning models such as deep learning/neural networks, random forest, boosting regression trees, and variable selection to predict excess market return.