Alessio MEZZINA

PhD Student
PhD in Computer Science - Ciclo 40°
Tutor: Mario Francesco PAVONE

Education

  • Ph.D. in Computer Science, University of Catania, ongoing (since 2024)
    • Semantic Generative AI (corporate-funded Ph.D. fellowship with Koexai srl)
  • Master’s Degree in Computer Science, University of Catania, 2024
    • Thesis: "Real-Time Adaptive Metaheuristic for Handling Dynamic Optimization Problem. A Case Study on Dynamic Map Labeling Problem"
    • Final Grade: 110/100 cum laude
  • Teaching Qualification, University of Catania, 2024
    • Subject Area: A041 – Computer Science and Technology
  • Bachelor’s Degree in Computer Science, University of Catania, 2022
    • Thesis: "Visual Sonar: Underwater Acoustics Simulator and Sonar Systems"
    • Final Grade: 110/110 cum laude

Work Experience

  • Maggio 2025 - Ottobre 2025, Junior Teaching Assistant, University of Catania
    • Algorithms and Lab course (Bachelor’s Degree in Computer Science, L-31 – DMI)
    • Operating System course (Bachelor’s Degree in Computer Science, L-31 – DMI)
  • February 2024 – July 2024, Junior Teaching Assistant, University of Catania
    • Algorithms and Lab course (Bachelor’s Degree in Computer Science, L-31 – DMI)
  • October 2023 – March 2024, Junior Teaching Assistant, University of Catania
    • Applied Informatics course (Bachelor’s Degree in Biological Sciences, L-13 – DIPBIOGEO)
  • February 2023 – December 2023, SDE Intern, STMicroelectronics
    • Developed proof of concepts for Power Automate Desktop, technical documentation, and support for the data science team
    • Created POCs and documentation for Power Automate Desktop; worked closely with an international team based in India

My research focuses on generative Artificial Intelligence, with particular emphasis on the integration of Large Language Models (LLMs), deep neural networks, and advanced optimization techniques to enhance large-scale model performance. His work also includes the application of metaheuristics to optimize computational processes, both in real-time contexts and in industrial or academic settings.

VIEW THE PUBLICATIONS
N.B. the number of publications can affect the loading time of the information