MSc Computer Science, AI
110/110 cum laude with special mention to the career. Thesis on hybrid graph-RAG for public procurement at SWAP.
Education
BSc Computer Science
110/110 cum laude. Thesis on aerial image segmentation at CILab.
Publications
Weather-Aware Multimodal Sensor Fusion for Robust Perception
Camera-radar fusion for vehicle detection in adverse weather, on the RADIATE dataset.
FFT-UniBa at Cruciverb-IT pdf code
Italian T5 with length tokens and a CSP grid solver for crosswords: MRR 0.63, 34% of grids fully solved.
Experience
Didactic Tutor
Programming and Formal Languages for first-year students.
Projects
Tender Copilot
Local graph-RAG over 103M+ procurement records.
Cuelight
Present a PDF together, every screen in sync.
Look Where It Matters
Grad-CAM-guided distillation of a VLM into a ViT.
Code Comment Classification
Multi-label classifier with a full MLOps pipeline.
Quantum Contrastive Learning
SimCLR with quantum projection heads.
Temporal Link Prediction
EvolveGCN on a social network.
Credit Rating Prediction
86% balanced accuracy on imbalanced data.
Awards
Leonardo Hackathon, Space
€2000 prize. Finalist out of 400.
Italian University AI Competition
Out of 20 teams.
ITAData Hack
Out of 15 teams.
AI2B Hackathon
Out of 10 teams, €500 prize.
CyberChallenge
Local stage. 21st of 43 teams at the national stage.
Languages
Italian, English (B2)