The bias your LLM forgets when you log-transform
Fit a regression on log Y, exponentiate the prediction back, and you've quietly introduced a Jensen-inequality bias. Every codebase I've seen ignores it — and LLMs cheerfully …
MSc in Statistics
2023-03
2026-03
University of São Paulo
BSc in Economics
2015-03
2019-12
University of São Paulo
I spend my time turning messy real-world data into useful information and actionable initiatives.
Lately that’s looked like building agentic AI for marketplace platforms, designing fraud and data-leakage detection across sensitive channels, and shipping causal and time-series models that ground business choices in something more than intuition.
Before tech, I led growth analytics for Latin America’s largest wholesale bank and ran economic research for São Paulo state.
I write occasional notes on this site. Reach out if any of that overlaps with what you’re working on.
Fit a regression on log Y, exponentiate the prediction back, and you've quietly introduced a Jensen-inequality bias. Every codebase I've seen ignores it — and LLMs cheerfully …