Machine learning pipelines, scientific simulations, and software I have built.
Mirion Technologies
Physics R&D InternBuilt C++ Geant4 Monte Carlo simulations and trained machine learning classifiers to identify radioisotopes from continuous Compton spectra without discrete photopeaks.
UConn Computational Galaxy Formation Lab
Astrophysics AI ResearcherUnderstanding CNNs trained to predict cosmological parameters directly from CAMELS hydrodynamic simulation volumes, using gradient attribution to trace what physical features drive network outputs.
UConn Computational Galaxy Formation Lab
Astrophysics ResearcherExtracted 3D filamentary networks from cosmological simulations to benchmark how baryonic feedback processes alter cosmic web topology.
helio
Co-FounderCollecting real-time rooftop solar irradiance data at UConn and training time-series deep learning models to predict localized power drops.
Wolfram Research
Ambassador & ResearcherAutomated image processing pipelines in Wolfram Language to calculate fractal box-counting dimensions of active solar flare regions.
Wolfram Summer Research Program
Student ResearcherProcessed raw spatial hit matrices from Timepix silicon pixel detectors to filter particle hits and evaluate radiation shielding.