English name: Stella
Shi Ding is an award-winning researcher at the Georgia Institute of Technology, where she explores how frontier AI and interactive agentic systems can scaffold human learning and agency rather than induce cognitive over-reliance. She is a Ph.D. Candidate in Digital Media at the Expressive Machinery Lab, advised by Regents’ Professor Brian Magerko. Her research broadly focuses on human–AI interaction, human-centered AI evaluation, and accessible learning technologies.
Ding’s research has received a Best Paper Honorable Mention at an ACM SIGCHI-sponsored conference (IDC 2026), the CRIDC Research Award, and an AI Child Safety Lab Fellowship. Her publications span Human-AI interaction, Human-centered AI, accessible computing, and AI literacy, including ACM IDC, SIGCSE, ASSETS, and AIED, alongside first-author contributions at NeurIPS and ACL workshops.
Beyond her core research, she has collaborated with diverse interdisciplinary faculty at Georgia Tech—including Mark Riedl, Richmond Wong, Chris MacLellan, Michael Nitsche, and Jason Freeman.
She has presented her research at Harvard University, Delft University of Technology, UC Davis, UC Irvine, and the Georgia Tech Responsible Computing, AI, & Society Summit.
As an active leader in the academic community, she has served as an Area Chair for ACM DIS, a Session Chair for ACL, and a reviewer for top-tier conferences including CHI, DIS, FAccT, AMIA, C&C, NeurIPS, SIGCSE TS, and CogSci.
In industry, she previously worked as a Research Scientist at the AI startup Grasply AI. She is also a CREATE-X Fellow and the co-founder of Deeproject.AI, supported by the Quadrant-i Startup Grant, with mentorship from Richard Gruber and William Shawn Yeager.
Designing and developing a context-aware LLM assistant for creative coding. The assistant combines conversation history, curriculum retrieval, live code state, and platform tools. My work connects co-design with secondary school learners and educators to assistant requirements, implementation, and evaluation.
Making music programming more accessible to blind and visually impaired learners through sustained co-design with learners and educators. My work focuses on platform accessibility, non-visual interaction, and translating user needs into design requirements within the NSF-funded EarSketch project.
Developing human-centered approaches to understanding and evaluating AI across multi-turn interactions.
Through Deeproject.AI, I explore AI evaluation across domains, including digital media, engineering, fintech, and healthcare.