Research
My research sits between machine learning and the communities it affects. One thread asks how local governments regulate AI in the absence of federal rules, and how data center development changes the places that host it. Another builds forecasting and representation-learning tools for energy and climate systems.
Draft text: replace with your research statement.
Social Algorithms Group (Prof. Karrie Karahalios) · 2025 – presentLocal AI governance and the community dynamics of data centers
Qualitative study of how U.S. localities regulate AI absent federal rules, and how data centers affect host communities.
- Collected all 77 AI policies from 55 U.S. local governments through systematic searches and policy repositories.
- Conducted the literature review, contributed to study design, and co-wrote the methodology.
- Began in the Social Spaces Lab at UIUC, which has since become the Social Algorithms Group.
IBM-Illinois Discovery Accelerator Institute · 2024 – 2025Representation fusion for climate foundation models
Team project extending IBM/NASA's Prithvi climate foundation model with graph-based and multimodal representations for weather, crop, and climate-risk forecasting.
- Co-developed, with Dhanish Natarajan, neural network models predicting crop indemnities across 3,144 U.S. counties; adding GraphCast representations reduced test MSE by 15% (R² 0.70 → 0.74).
- Contributed to design discussions for a multimodal architecture fusing satellite imagery with weather-station data across spatial and temporal scales.
- Contributed to design discussions for extending Prithvi with graph-based representations; the team's framework improved masked-region and autoregressive prediction.
- Presented at the UIUC Undergraduate Research Symposium (2025).
- PIs: Dr. Jingrui He, Dr. Hanghang Tong, and Dr. Elizabeth Ainsworth. Supported by the Illinois Scholars Undergraduate Research (ISUR) program.
IIDAI URE FellowContextual Engineering Research Group, UIUC · 2023Contextual engineering for community water systems
Qualitative and computational methods to center community needs in the design of water and power infrastructure.
- Co-developed, with Annabelle Epplin and Cathy Li, a qualitative coding process and master codebook for community interviews about a water system in Clarinero, Honduras, reaching moderate-to-substantial inter-rater agreement (Fleiss' κ = 0.56–0.69).
- Modeled community characteristics via entropy maximization to inform water infrastructure design.
- REU fellowship in Bodaway Gap, Arizona, on the Navajo Nation, studying ethnography in the algorithm design process.
- Presented at the UIUC Undergraduate Research Symposium (2023).
Outstanding Research Poster AwardFlorida Atlantic University (Dr. Zhen Ni) · 2021 – 2022Federated learning for smart grids
Federated learning approaches for forecasting in carbon-neutral smart homes and grids.
- Presented at the FAU OURI Undergraduate Research Symposium (2022).
