I'm a master's candidate at the UC Berkeley School of Information, researching natural language processing, with a focus on robust evaluation methods for large language models. My current work develops perturbation-based frameworks that better predict how models perform under real-world conditions, which adds the missing piece to the gap between benchmark performance and actual deployment reliability.

At UC Berkeley, my focus is on computational linguistics, neuroscience applications of NLP, and machine learning for remote sensing. I'e conducted computational analysis of 74,000+ film dialogue lines to identify systematic linguistic differences between genres, using speech act detection, named entity recognition, and conversation structure analysis to build interpretable classification models. In collaborative work on language and cognition, I compared Word2Vec, GloVe, and BERT embeddings for predicting voxel-wise fMRI brain activity from narrative text, implementing preprocessing pipelines for temporal alignment and hemodynamic delay modeling. I've also developed hybrid feature engineering approaches combining domain-informed spatial features with autoencoder-learned representations for cloud detection in Arctic satellite imagery. For my capstone, I'm working with a team to build a tool for systematically evaluating large language model robustness, investigating how input variations affect model behavior and reliability.

Before graduate school, I workedas a software engineer in Nairobi, Kenya. At Flux Water Limited, I designed and optimized backend systems using Django and PostgreSQL. I built real-time operational monitoring dashboards for water supply infrastructure and created AccessWASH, a nationwide water-access data platform handling over 100,000 datapoints across microservices, serving thousands of users advocating for better water, sanitation, and hygiene services across Kenya. That engineering background informs how I approach research with attention to reproducibility, efficiency, and real-world impact.

I'm a recipient of the Mastercard Foundation Scholarship, the I School Fellowship, and the Dr. James R. Chen Award.