Computer Science student at Politecnico di Milano.
I build research systems, data-heavy tools, and full-stack applications.
I am interested in the intersection of:
- machine learning systems,
- quantitative research infrastructure,
- data validation and reproducibility,
- high-performance computing,
- full-stack products that real people can use.
My current focus is learning how to build software that does not just produce nice-looking results, but can also explain where those results came from and why they should or should not be trusted.
A research infrastructure for falsifiable equity-strategy experiments.
This is not a trading bot and not a promoted strategy. It is a lab for testing whether strategy ideas survive data-quality checks, transaction costs, out-of-sample validation, robustness gates, provider constraints, and an audit trail.
The project includes:
- governed pre-run gates before provider queries or backtests,
- decision ledgers and vault reports for every research outcome,
- outlier-dependency and cost-realism gates,
- regime-aware portfolio diagnostics,
- Streamlit dashboards for inspecting strategies, blockers, and results,
- a frozen monthly replication protocol for the remaining active hypothesis.
Main lesson: the lab has promoted zero strategies so far, which is exactly the point. It is built to reject weak evidence before it becomes a trading claim.
Python | pandas | scikit-learn | Streamlit | time series |
quantitative research
Full-stack web application for market stall booking and event management, built for Pro Loco Soresina and deployed in production.
Next.js | Supabase | TypeScript | PostgreSQL
Computer-vision research prototype for detecting cephalometric landmarks from radiographic/profile images.
C# | EmguCV | OpenCV | image processing
Multiplayer chess application with Java RMI synchronization, legal move highlighting, turn handling, check/checkmate detection, and castling support.
Java | Swing | RMI
Java Swing typing trainer with accounts, timed challenges, scoring, difficulty modes, and leaderboard screens over a socket-based protocol.
Java | Swing | sockets
Languages: Python, Java, TypeScript, JavaScript, C#, C
Data / ML: pandas, scikit-learn, time-series validation, Streamlit
Web: Next.js, Supabase, PostgreSQL, REST APIs
Systems: Git, testing, experiment logging, reproducible pipelines
I am using my first years of university to build a foundation in computer science while turning serious personal projects into public case studies.
The long-term goal is to work on technically demanding systems: research infrastructure, ML tooling, high-performance software, and products where correctness matters.
Cremona, Italy | Politecnico di Milano