Skip to main content

About me

Background

I am from Juiz de Fora, Brazil, and my education started in Physics at the Federal University of Juiz de Fora (UFJF). During my Master’s in Physics (Materials Science) and as a researcher in applied physics, I worked on modeling complex systems and quantitative analysis of experimental data in Python: uncertainty propagation, validating models against measurements and communicating technical results.

My postgraduate degree in Data Science completed the transition. Since then I have built complete projects, from raw data to deployed apps, with tabular data, time series and Portuguese text. I follow the same method in all of them: an explicit baseline before any complex model, a metric chosen for the problem (PR-AUC under imbalance, QWK for ordinal targets, RMSPE for forecasting) and SHAP to explain what the model learned.

Experience

Role Where Period
Data Scientist — independent projects (12 published projects) Personal portfolio · Remote Apr 2026 – present
Applied Physics Researcher Federal University of Juiz de Fora Apr 2022 – Sep 2024

Education

Degree Institution Period
Postgraduate degree in Data Science UNINASSAU 2025
M.Sc. in Physics (Materials Science) Federal University of Juiz de Fora 2022 – 2024
B.Sc. in Physics Federal University of Juiz de Fora 2019 – 2022

Certifications: Ciência de Dados Impressionador (Hashtag Treinamentos, 2026) · Algorithms and Programming Logic (Hashtag Treinamentos, 2026)

Technical skills

Languages

Python SQL CTEs Window functions

Modeling

scikit-learn XGBoost LightGBM statsmodels Prophet PyTorch BERTimbau

Interpretability & tuning

SHAP Optuna Ray Tune Hypothesis testing

Visualization & BI

Power BI Matplotlib Seaborn Plotly GeoPandas Folium

Delivery

Streamlit Git/GitHub JupyterLab conda SQLite

Familiar with

R AWS (fundamentals)

What I am looking for

I want to work as a Junior Data Scientist or Data Analyst, in teams that value evidence-based decisions, where I can turn technical analyses into practical business recommendations.

Back to top