|
I hold a B.Sc. in Computer Engineering from the
Federal University of São Carlos (UFSCar), Brazil.
I previously worked as a Visiting Researcher at Linköping University (Sweden), developing AI-based control systems for autonomous UAVs. That experience reinforced my interest in combining rigorous modeling with practical engineering. Today, I work as an AI Engineer at VLS Legal Capital, at the intersection of Machine Learning, NLP / Document AI, Advanced RAG, and Business. Core themes: Document AI & NLP • RAG • Applied ML • Analytics for Business |
Languages & Data
ML / Data Stack
Engineering & Tools
Robotics
BI / Analytics
- Machine Learning & Applied AI Engineering
- NLP / Document AI (OCR, extraction, classification)
- Retrieval, RAG, reranking, evaluation
- Business analytics, decision support, and strategy
- Robotics perception & autonomy (ROS2)
-
Telco Customer Churn Prediction
End-to-end classification pipeline with SMOTE, cross-validation, and multiple ML models (SVM, MLP, Naive Bayes). -
Market Basket Analysis
Frequent Itemset Mining + Decision Tree on Online Retail II to discover purchase patterns and translate them into actionable insights (Apriori + classification).
-
PathPursuit
Mobile app for autonomous robot navigation using ROS and JavaScript. -
Matrice100_dev
ROS-based framework for path planning and adaptive control for DJI Matrice 100 UAVs.
-
G. Bertho, R. S. Inoue, M. Wzorek, P. Rudol,
Deep Neural Network-Based LQR Adaptive Control for Commercial Quadrotors Using ROS,
2025 Brazilian Conference on Robotics (CROS), Belo Horizonte, Brazil, 2025, pp. 1–6.
DOI: 10.1109/CROS66186.2025.11066163 · IEEE Xplore · DOI -
B. Yui, G. Bertho, K. Barcelos, R. Yamauchi, H. Guardia,
Uma Abordagem para o Knight’s Tour Problem utilizando Depth First Search,
Anais da XVI Escola Regional de Alto Desempenho de São Paulo (ERAD-SP), São José do Rio Preto/SP, 2025, pp. 54–57.
DOI: 10.5753/eradsp.2025.9755 · SBC OpenLib (SOL) · DOI
If you’d like to collaborate or discuss AI, data products, or robotics, feel free to connect:
