Projects
ML/Data Science for Sparse Ocean Warming Data Analysis
Tech: Python, PyTorch, AWS (S3, EC2), Docker, Coiled (Dask), Xarray, NetCDF
Developing a probabilistic 3D field reconstruction system using physics-informed ML to interpolate sparse Argo float data into uncertainty-quantified subsurface temperature profiles. Building scalable AWS/Docker inference pipelines for 20-year multi-terabyte geophysical datasets.
In progress — Dec 2025–
ML Aquaculture Risk Estimation
Tech: Python, PyTorch, CNN, Transformer, GCP
Building a deep learning risk model predicting threshold-crossing heatwave probabilities with calibrated probabilistic outputs. Engineering GCP data ingestion and feature pipelines for end-to-end model training and serving.
In progress — Mar 2026–