BSc in Computer Science, 110/110 with Honours
Thesis on the reproducibility of recommender systems, focusing on aligning evaluation metrics within the ClayRS open-source framework.
Posts by Collection
education
MSc in Data Science
An advanced master’s program focused on state-of-the-art techniques in machine learning and statistical analysis, pursued while working full-time as a Data Scientist at IBM.
hackathons
ITADATAHACK 2024
Competed in a 72-hour hackathon on a time series problem as part of a team of 3, achieving an F1 score of 70%.
VisionAI Assistant (LauzHack 2024)
A multi-modal AI assistant developed in 22 hours, capable of analyzing images using natural language voice commands.
portfolio
Smart Traffic Lights (Knowledge Engineering Project)
An AI-powered system that optimizes traffic flow by using a Prolog knowledge base, A* search, and HMMs for traffic prediction.
AquaAware: A Two-Phase Learning System for Water Conservation
A smart water-saving system that uses a two-phase machine learning approach to train users and optimize household water consumption.
Baldness Classification Project
A deep learning project to classify different stages of baldness using CNNs and pre-trained models like ResNet50 and VGG16.
Dark Web Topic Analysis (GSoC @ HumanAI)
A project focused on uncovering how language tied to criminal behavior evolves over time within Dark Web forums.
Blog PostVisionAI Assistant (LauzHack 2024)
A multi-modal AI assistant developed in 22 hours at EPFL, capable of analyzing images using natural language voice commands.
Gemma Model Benchmark Suite
An open-source toolkit designed to democratize LLM analysis and benchmarking with deep customization and minimal compute requirements.
Blog Postpublications
teaching
volunteering
Mentee at Superhero Valley
Selected as one of the Italian students for the mentorship program of Superhero Valley, a community dedicated to guiding students in computer science fields.
work
Data Scientist AI Associate
Designing and deploying end-to-end AI solutions, focusing on NLP, MLOps, and large-scale performance optimization on IBM Cloud.
Open Source Contributor
Analyzed 500k Dark Web discussion points using NLP models to identify key topics and deployed predictive models on Hugging Face to forecast trends.