Ing. Marcelo Bozunovsky

Multidisciplinary engineer with formal training in electronics, computer engineering, and applied mathematics, and more than fifteen years of experience in the design of complex software systems, data platforms, and artificial intelligence solutions.

PhD candidate in Biomedical Engineering at the University of Buenos Aires, focused on computer-aided diagnosis (CAD) systems based on computer vision for medical imaging (specialty: Breast Cancer).
I investigate novel methodologies, in collaboration with Prof. Dr. Ing. Silvano Zanutto, aimed at developing clinically relevant CAD tools for digital breast tomosynthesis.

Physics Department Professor. Areas: Mechanics, Optics, Electricity & Magnetism (Physics II). Other areas: Probability.

• Universidad de Buenos Aires, Ingeniería en Informática 2015-2025 (MSc in Computer Science)
• Universidad de Buenos Aires, Ingeniería en Electrónica 2015-2024 (MSc in Electrical Engineering)
• Universidad de Buenos Aires, Maestría en Ingeniería Matemática 2024-2025 (MSc in Mathematical Engineering)
• PhD Candidate in Biomedical Engineering

• Artificial Intelligence & Deep Learning. Design, training, and analysis of convolutional neural networks with a strong focus on ResNet-style architectures, residual connections, and deep feature hierarchies. Extensive experience with patch-based and full-image models, transfer learning, fine-tuning strategies, and evaluation metrics such as AUC, ROC curves, and mAP in real-world settings.
• Signal Processing Foundations. Solid background in digital signal processing, including filtering, spectral analysis, time–frequency representations, convolution operators, noise modeling, and statistical feature extraction. Practical application of DSP concepts to image formation, preprocessing, normalization, and robustness analysis in machine learning pipelines.
• Computer Vision & Medical Imaging. Development of computer vision systems for medical image analysis, including mammography and other diagnostic modalities. Experience with image enhancement, segmentation, multi-scale analysis, and CNN-based feature learning for computer-aided diagnosis (CAD) systems.
• Large-Scale Data Platforms & Analytics. Architecture and implementation of high-volume data platforms, including ETL pipelines, analytical data models, and performance-oriented database design. Integration of machine learning workflows with production-grade data infrastructures.
• Mission-Critical & High-Availability Systems. Design of robust, fault-tolerant systems with strict reliability and performance requirements. Experience with monitoring, logging, error handling, and operational resilience in production environments.
• End-to-End System Design. Full lifecycle ownership, from mathematical modeling and algorithmic research, through prototyping and validation, to deployment and long-term maintenance of production systems that combine software engineering, data engineering, and AI.

• Early medical diagnosis through artificial intelligence. My MSc. Electronics Engineering Thesis: Desarrollo de un sistema CAD para análisis de mamografías usando redes neuronales convolucionales y redes DBN.
• Accessibility and assistive technologies
• Automation of complex, knowledge-intensive processes
• System scalability, robustness, and reliability
• Engineering solutions with real-world impact, using real-world demanding datasets
• Baby apnea detection - SMSL