I'm a Research Scientist & Engineer focused on building and validating Machine Learning systems in highly demanding environments. My background merges rigorous quantitative methods with high-performance computing (HPC) to solve complex data challenges—from real-time signal processing at CERN (40M events/s) to industrial anomaly detection.

Currently, I'm a Master's student in Artificial Intelligence at UFRJ, applying deep learning to critical electronic instrumentation.

• Open Source Contributor (Petrobras 3W Toolkit): Building evaluation modules for the 3W Toolkit, an open-source tool for the Oil & Gas sector. My focus is on developing benchmarks to assess ML model robustness, reproducibility, and reliability in critical operational scenarios.
• Research Scientist @ Fundação COPPETEC: Engineering scalable Deep Learning pipelines in collaboration with Petrobras to predict complex physical properties.
• Research Scientist @ Signal Processing Laboratory/UFRJ: Developing Deep Learning models for Anomaly Detection and classification of passive sonar signals in collaboration with the Brazilian Navy.

• Model Risk & Robustness: Validating ML systems to ensure they perform reliably under uncertainty and stress.
• High-Performance Computing (HPC): Engineering data pipelines and real-time processing systems using C++ and CUDA.
• Scientific Machine Learning: Applying rigorous statistical modeling to physical systems.
• Signal Processing: Using DSP for real-time data correction, filtering, and quality assurance of sensor data.

• Languages: Python, C/C++, SQL, R, Bash
• ML/DL Frameworks: PyTorch, TensorFlow, Scikit-learn
• HPC & MLOps: AWS, Docker, Kubernetes, Slurm, Singularity
• Data & Databases: PostgreSQL, SQLite, MySQL, MongoDB, Pandas, NumPy

• Email: pedrohblisboa@gmail.com
• LinkedIn: /in/pbragali