Data Scientist and Machine Learning Engineer with over 4 years of experience developing scalable, data-driven solutions. Currently pursuing a Ph.D. in Data Science at PUC-Rio, focusing on scalable generative AI architectures and anomaly detection.

🎓 Academic Background:

• Approved in both IME and ITA entrance exams
• Gold medal winner at Desafio PUC-Rio Olympiad (full scholarship recipient)
• Master's in Applied Mathematics from PUC-Rio
• Bachelor's in Mathematics from PUC-Rio

🏢 Industry Experience:

• Worked on high-impact AI initiatives with Petrobras, Intel, Embraer, and Eletrobras
• Delivered solutions in: Generative AI, RAG, Computer Vision, LLMs, NLP, and Anomaly Detection
• Full ML lifecycle expertise: modeling, deployment, CI/CD pipelines, MLOps, cloud-native infrastructure

Machine Learning Staff Researcher and Engineer @ HVAR (Sep 2025 - Present)

• Research and development on LLMs, RAG, and GraphRAG solutions
• Working with LLAMA and GEMINI for text-to-SQL applications
• Designing advanced ML solutions within Databricks

• Deep Learning: Neural Networks, CNNs, RNNs, Transformers
• Classical ML: Random Forest, XGBoost, SVM, Decision Trees, KMeans
• NLP & LLMs: LLAMA, Gemini, GPT, RAG, Text-to-SQL
• Computer Vision: Object Detection, Image Classification
• Anomaly Detection: Isolation Forest, SOS, LOF

• Data Processing: Pandas, NumPy, Elasticsearch
• Visualization: Matplotlib, Streamlit
• Version Control: Git, GitHub Actions
• Deployment: BentoML, Flask
• Other: Unity Catalog, Unix Systems

• Developed anomaly detection system for petroleum well time series
• Achieved 80%+ detection rate using SOS, Isolation Forest, and LOF
• Enabled proactive maintenance and optimized resource allocation

• Built experimental RAG framework for SQL query generation
• Leveraged LLAMA, Gemini, and GPT architectures
• Automated natural language to SQL conversion

• Implemented ML-based fraud detection system
• Achieved 4x higher accuracy compared to manual methods
• Utilized clustering, regression, and decision trees

• Built end-to-end ML pipeline for default risk prediction
• Reduced debt by 50%+ and recovered $1M in losses
• Automated data pipelines with AWS and PySpark

Ph.D. in Data Science (2023 - 2026)
Pontifícia Universidade Católica do Rio de Janeiro (PUC-Rio)

M.Sc. in Applied Mathematics (2021 - 2022)
PUC-Rio | Thesis: Random Forest for Reservoir Simulation

B.Sc. in Mathematics (2018 - 2020)
PUC-Rio

Telecommunications Engineering (2015 - 2017)
Instituto Militar de Engenharia (IME)

• Portuguese: Native
• English: Advanced

I'm always interested in collaborating on innovative ML and AI projects. Feel free to reach out!

⭐️ From igorconsulting