I see Statistics as a way to understand phenomena: finding patterns, quantifying uncertainty and building models that represent real problems — that is the pretty explanation. The real one is that statistics and computing are my favourite tools for finding solutions to problems I enjoy.
Who am I?
I'm an undergraduate Statistics student at UFES, focused on Data Science, Machine Learning, AI and Data Engineering, combining statistical rigor with software engineering.
I like to think in products, end to end: requirements, users, weaknesses and real-world use — everything on paper before any code (even if I don't build it all on my own!). Curiosity, consistency and intensity (my defining trait) drive what I build. Not tool demos: projects that test hypotheses, expose their limits and actually solve something.
This is a meeting point of math, data and technology — projects, experiments and studies that track how I'm evolving.
I also keep here the posts, projects and texts I now disagree with, have already updated or am still updating; this is just a brief summary of what I build.
Formal training in Statistics, complemented by courses on taking machine learning to production.
Universidade Federal do Espírito Santo (UFES)
Data Science Academy
Data Science Academy
Data Science Academy
Data Science Academy
Research intern
Management and Data Consulting
Projects Director · CONSTAT · President · CONSTAT · Development Director · JUNIORES
Inference, uncertainty and model evaluation — calibration, resampling, decision thresholds. Mostly in R, also in Python.
→ articles02From baselines to tuned ensembles, with honest validation splits, interpretability and a decision rule chosen on purpose.
→ Credit Default API03Layered pipelines, columnar storage, cell-level lineage and data-quality rules that never fix data silently.
→ EduData · pipeline04Local models behind strict JSON contracts: the model interprets, code computes — and its accuracy is measured, not assumed.
→ EduData · llm evaluation05Requirements, users, limitations and real use — written down before the code, stated openly after it.
→ every project pagePython and SQL items appear in the code of the projects on this site; R is my day-to-day language for statistics.
Open to internships, research and data roles. Email is the fastest way.