Jacomo Corrieri

About Me

I’m a software developer, researcher, musician, and lifelong learner with a love for computing, math, and more. From conducting undergraduate research to studying abroad in Seoul, South Korea, I've always pushed myself to try new experiences. I bring a strong work ethic, experience tackling challenging problems, and diverse programming skills, primarily in research, machine learning, and full-stack application development.

pdfView my Résumé fine-printStudy Blog

Education

UF Logo Light

University of Florida

M.S. in Computer Science

Currently Enrolled

UNF Seal

University of North Florida

B.S. in Computing and Information Sciences – Computer Science

Minors in Mathematics, International Business

GPA: 3.99 / 4.0

Graduated summa cum laude; Member of Hicks Honors College

Skills

python--v1 Python
java-coffee-cup-logo--v1 Java
c-programming C Language
html-5--v1 HTML
css3 CSS
pytorch PyTorch
assembly Assembly
console Unix
git Git

Projects

Technologies Used

Multi-Robot Exploration with Deep Reinforcement Learning

This is a Deep Reinforcement Learning (DRL) framework for multi-robot coverage using Proximal Policy Optimization (PPO) with a centralized critic and decentralized actors (CTDE framework), which maintains connectivity during exploration. Following this framework, each robot learns a local policy based on partial observations, such as LiDAR scans, the visited cell history, and the states of nearby robots, while benefiting from centralized value estimation during training. The combination of convolutional feature extraction, centralized value estimation, and communication-aware reward shaping enables consistently high coverage and connectivity among multiple robots. This approach scales effectively with team size: we demonstrated successful training with up to 20 robots, and thanks to the decentralized policy design, the learned models generalize seamlessly to larger teams, achieving reliable performance with up to 50 robots at test time.

github View on GitHub

Technologies Used

  • FastAPI (Python)
  • Next.js / React (TypeScript)
  • TailwindCSS
  • Vercel / V0
  • GitHub (Team Collaboration)
  • External APIs for images, text, and Google Flights

Aeroatlas Travel Planner – Audience Choice Award Winner

We developed a travel itinerary planner that generates custom trip plans with activities, images, and local insights based on user input. The project was awarded the Audience Choice Award at the School of Computing Symposium at UNF. As part of a five-person team, I focused on developing the dynamic itinerary page, integrating a FastAPI backend with several APIs to fetch images (flags, attraction previews), trip descriptions, and user-saved itinerary data. Users can add restaurants, attractions, hotels, and custom items to their itinerary, manage a packing list, and view flight information via Google Flights. All changes persist on the backend to support a fully interactive experience.

github View on GitHub

Technologies Used

  • Java
  • JavaFX & CSS (GUI Development)
  • Linux
  • SSHJ - SSHv2 library for Java

Raspberry Pi Interface

This project implements a desktop interface in Java to control a Raspberry Pi system. It features a graphical UI to interact with the Pi’s components and monitor key system resources. Key modules include a file manager for uploading, downloading, renaming, and deleting files; a basic shell interface; and real-time CPU, RAM, disk usage monitoring, and a GPIO control panel (currently read-only). The application makes use of JavaFX Tasks to handle multithreaded operations, ensuring background processes such as loading and data collection do not block program usage. Users can save their system preferences for greater convenience when re-launching the application.

github View on GitHub
View All (+)

Contact