Alejandro Mezio

I'm a physicist (PhD in condensed-matter physics) who moved into data science, bringing 15+ years of quantitative and statistical modeling from research into applied problems. My focus is turning ambiguous questions into working models, forecasting, predictive analytics, time series, and statistical inference, with a strong emphasis on rigorous evaluation, honest uncertainty, and avoiding data leakage.

I work primarily in Python, and I care about interpretable models and reproducible, production-style pipelines. Recent portfolio work includes a delivery-time forecasting system with temporal (leakage-aware) validation, and a leakage-audited classification case study.

Alongside research, I've taught university courses in data science, mathematics, and physics, which sharpened how I communicate technical results to non-technical audiences.

I'm looking for remote data science and ML roles (I'm based in Argentina, UTC-3), especially in forecasting, fintech and credit risk, energy and climate time series, or marketing analytics.