I'm a freshman at UC Berkeley studying Data Science through CDSS. Before Berkeley, I was a Regeneron ISEF finalist and published two papers on triboelectric nanogenerators — work that taught me how much ground you can cover when curiosity outpaces the syllabus.
Engineered nanocomposite materials with multiwalled carbon nanotubes to enhance triboelectric charge density, with durability analysis under repeated mechanical loading.
Designed a self-powered sensing architecture using water-contact electrification, optimizing voltage stability under dynamic environmental conditions.
Developed a hybrid renewable energy framework integrating triboelectric generators with conventional sources — a scalable model for distributed energy resilience.
An AI app for long-form video understanding. Transcription, automatic chapter generation, semantic search across footage, and a clip-builder for extracting moments. Python/FastAPI backend running vLLM on RunPod, SwiftUI frontend.
iOS app for tracking orthodontic alignment progress over time. Built with SwiftUI, Firebase Authentication, and Firestore for real-time sync across devices.
Deep convolutional architecture for medical image classification on chest X-ray datasets, with preprocessing pipelines and cross-validation to improve generalization.
Recurrent neural network models for multivariate time-series forecasting, evaluated with RMSE and comparative loss convergence analysis.