Mateus holds a PhD from École Centrale de Lyon and RMIT University, where his research focused on photonic convolutional neural networks and their potential for energy-efficient AI hardware. His doctoral research investigated the impact of hardware non-idealities on photonic neural network performance, including quantization and noise effects, as well as the trade-offs between computational precision, classification accuracy and energy efficiency. The work also explored strategies for improving the utilization of photonic computing resources while maintaining hardware efficiency and scalability.
Mateus also holds a degree in Physics Engineering from UFRGS, with a strong background in mathematics, physics and engineering. During his undergraduate studies, he worked for nearly two years on superhydrophobic nanoparticles and developed a robust RPA algorithm for XAS deglitching as his final thesis project. He also has two years of professional experience as a data scientist, combining scientific research, computational methods and data-driven problem solving.
LinkedIn: linkedin.com/in/mvidaletticosta
