Gabriel Estrela Lopes
I am a data scientist with a solid statistical foundation and hands-on experience in predictive modeling (machine learning, deep learning) and mathematical optimization.
About
I am a data scientist with a solid statistical foundation and hands-on experience in predictive modeling (machine learning, deep learning) and mathematical optimization. I specialize in building AI architectures, with a recent focus on LLMs and retrieval-augmented generation for conversational solutions.
Currently an AI & Advanced Analytics Intern at Grupo Energisa, I improved credit risk scoring with XGBoost and logistic regression (+10% accuracy), engineered large-scale data pipelines with SQL/Spark, and developed MILP optimization for fleet routing (+20% logistic efficiency). I also automated data handling with Python-based RPA and supported teaching in statistics and probability at UFPB.
- 1 year's experience
- Joined March 2026
Languages
- EnglishAdvanced
Industry experience
- Energy & Utilities
- Financial Services
- Software & Internet
- Education
- Professional Services
Work experience
Predictive & Credit Risk Modeling: Engineered supervised ML classification models (XGBoost, Logistic Regression) for financial default scoring. Applied SMOTE and hyperparameter tuning, resulting in a 10% accuracy increase in detecting high-risk profiles within the credit market. Data Engineering & SQL: Architected massive data pipelines using SQL, Apache Spark, and PySpark to process both structured and unstructured datasets. Applied clustering to compress strategic database volumes by 90%, optimizing model ingestion processes. Operations Research: Developed MILP algorithms in Python for optimized fleet routing, boosting logistic efficiency by 20%. Automation: Built Python-based RPA bots, saving 2 man-hours daily and eliminating manual errors in data handling.
Delivered practical sessions in Descriptive Statistics, Inference, and Hypothesis Testing (A/B testing), raising pass rates by 50%.
Education
Hire Gabriel Estrela Lopes today
Post a job and invite Gabriel, or contact them directly. No hidden commission.