Summary
Maria De Castro is a theoretical physicist-turned-data scientist with 9 years of experience building statistical and simulation-based models that separate signal from noise in complex, real-world systems. She blends deep expertise in uncertainty quantification, conformal prediction, and time-series Transformers with practical engineering—coding in Python/PyTorch and producing training materials and reproducible notebooks for diverse audiences. Maria teaches AI and programming at university level, develops workshops on Trustworthy AI, and has implemented compute-to-data services for petabyte-scale climate datasets, combining research rigor with user-focused documentation and support. Her background spans academia, climate and health data science, and science communication, and she notably led outstanding projects on tumor diagnosis from perfusion MRI and AI for sustainable energy transition. Colleagues describe her as a translator between theory and practice who makes probabilistic thinking accessible to engineers, students, and policy stakeholders.
10 years of coding experience