
Computer Engineer & AI Researcher
Computer Engineering graduate from Politecnico di Torino in June and incoming M.Sc. candidate in Computer Science with an AI focus at KU Leuven. Research Assistant at Politecnico di Torino, with experience in AI systems, full-stack development, and embedded firmware engineering.
Building the future through code and curiosity
KU Leuven
Politecnico di Torino
Technologies I work with daily
Practicing algorithmic problem-solving on LeetCode to sharpen my technical skills.
problems
acceptance
View my full profile on LeetCode
Building products and contributing to innovative teams
Politecnico di Torino – Prof. Paolo Montuschi
Designing and implementing machine learning models to analyze and predict student academic performance. Building scalable data pipelines and evaluation frameworks for large educational datasets. Designing and deploying AI-based class support systems integrated within university infrastructure.
Parma Calcio 1913
Built and deployed 20+ production full-stack features (React, TypeScript) used by 50+ staff, reducing operational workflow time by 50%. Integrated AWS S3, CDN delivery, and real-time data pipelines; achieved Lighthouse performance scores consistently above 95.
PoliTo Rocket Team
Developed internal web platforms for 150+ members, automating 1,000+ daily operations and reducing administrative overhead by 95%. Implemented STM32 flight firmware in C/C++ under real-time constraints; built drivers, watchdog systems, and performed SWD debugging.
Team DigiTwin – PoliTo
Designed and optimized PostgreSQL systems ingesting 50,000+ sensor readings/day; improved query performance by 40%. Implemented automated data pipeline processing 2GB+ of telemetry data weekly for structural monitoring.
Crafting solutions from concept to deployment
GSoC 2026 Applicant — ML4SCI / DeepLense
Completed evaluation tests for Google Summer of Code 2026 under ML4SCI / DeepLense. Multi-class substructure classification: AUC 0.9919, surpassing the Varma et al. reported baseline on the same dataset. Lens finding under 100:1 class imbalance: AUC 0.9877, 189 of 195 lenses recovered. Proposed Phase 2 applies domain-adversarial adaptation to close the sim-to-real gap on real HSC-SSP survey data.
MALTO Hackathon
6-class classifier distinguishing human writing from 5 LLM sources (DeepSeek, Grok, Claude, Gemini, ChatGPT) on a heavily imbalanced dataset. Ensemble of 3 TF-IDF models — word n-grams, char n-grams, and 20 handcrafted stylometric features — with meta-learner stacking. Top 10 finish out of all participants.
Personal Project
(WIP) Bird's-eye view strategy game where you build a timber empire from the ground up. Command workers to cut forests, process wood into refined materials, automate production chains, and scale operations. Features resource management, worker AI, and progression systems built with Java and JavaFX.
Open Source
(WIP) Privacy-first macOS system monitor using ML-powered anomaly detection. Unlike traditional monitors with hard-coded thresholds, CoreMetric uses a Reconstruction Autoencoder running on Apple Neural Engine to learn your usage patterns and detect subtle anomalies—memory leaks, crypto-miners, frozen processes—with <1% CPU overhead.
Open Source
(WIP) AI-powered study companion: ingest PDFs & notes → summaries, flashcards, interactive Q&A and milestone-based study schedules. Incremental AI pipeline with secure per-user isolation.
Roadmap Progress
29%Open Source
Production-grade Git simulation engine: dry-run rebase, merge, reset, cherry-pick with visual commit graphs, conflict prediction (3 certainty levels), and safety analysis. Awarded 3rd Place at GitKon Game Jam 2025. 135+ tests, automated CI/CD, interactive TUI, plugin architecture. Available on PyPI: pipx install gitsimulator (v1.0.1).
Personal Project
Modern portfolio website built with Next.js 16 and TypeScript, achieving 95+ Lighthouse score across all metrics. Implements server-side rendering, optimized image loading, and accessibility best practices (WCAG AA). Features dark mode, responsive design, and form validation with 99.9% uptime on Vercel.
Parma Calcio 1913
Architected and deployed 5+ full-stack features for Parma Calcio's internal operations platform serving 50+ staff members. Reduced data entry time by 35% through automated workflows using Next.js and CloudFront + S3, serving edge-cached documents and training videos to enable faster teamwide communication and data sharing. Implemented responsive UI components with 95+ Lighthouse performance score.
Team DigiTwin
Engineered high-performance database management system handling 50,000+ sensor readings per day for structural monitoring. Optimized PostgreSQL queries achieving 40% faster response times. Implemented C++ data processing pipeline with CMake for cross-platform deployment across Linux and embedded systems.
PoliTo Organization
Automated role assignment system for 200+ member university Discord server, reducing manual role management time from 2 hours/week to 5 minutes/month. Built with Discord.js and Node.js, integrated with Supabase for real-time data synchronization. Handles 1000+ role assignments with 99.9% uptime.
Open Source
Real-time WebSocket chat application built with Swift NIO for high-performance asynchronous networking. Features non-blocking I/O, concurrent connection handling, and event-driven architecture. Demonstrates modern Swift server-side development with efficient resource utilization.
Personal Project
Feature-rich desktop notes application with clean JavaFX interface for organizing and managing personal notes. Includes note categorization, search functionality, and persistent storage. Built with modern Java practices and MVC architecture.
Exploring AI, education technology, and computer architecture
Actively exploring research opportunities in the fields of Computer Science, AI, Architecture and more. Open to collaborations and academic partnerships.
Some of my recent Awards and Achievements
Awarded 3rd place at GitKon Game Jam for GitSimulator, a production-grade Git simulation engine. Recognized for innovative approach to Git visualization, educational impact, and focus on preventing costly mistakes in production environments.
View announcement →Awarded the Programme Scholarship with a 70% tuition waiver for the M.Sc. in Computer Science (Artificial Intelligence). Granted to only 10 students across the programme, recognizing academic excellence and research potential.
Awarded full-ride merit scholarship (90% GPA) for outstanding academic excellence in Computer Engineering. Covers complete tuition and living expenses at one of Europe's top technical universities for engineering.
Competed in Reply's European-wide Student Clash focused on Agentic AI, joined by 700+ students from 8 top universities across Europe. Designed and scaled innovative AI Agent projects that transform business operations and empower society through autonomous, intelligent systems.
Selected for exclusive hands-on datacenter hackathon hosted by Google Cloud in Turin. Worked directly with production-grade hardware including servers, managed switches, fiber optics, and Linux systems. Competed in team-based challenges simulating real-world datacenter operations, supported by Google's datacenter engineers from Turin and Milan facilities.