Education & Experience

Research Assistant at Bocconi University / Jul 2026 - Present.

Analyzing whether merger-related documents predict litigation involving target companies within one year of deal closing, using hierarchical Transformers with RoPE to pool SEC text into deal representations and combine them with quantitative FFNs through joint fusion and score ensembles; follow the work-in-progress research. Previously analyzed 10-K readability with topic, dispersion, classification, and embedding methods. Supervised by Professor Claudia Imperatore.

Generative AI Engineer Intern at Accenture / Jul 2026 - Sep 2026.

Developed an enterprise AI concierge using an LLM to interpret questions and route requests among models in an iterative loop designed to minimize latency between model requests. Built a Neo4j knowledge graph and GraphRAG pipeline combining entity relationships with semantic vector and keyword retrieval, achieving a six-percentage-point improvement over the existing RAG pipeline as measured by an LLM evaluator across approximately 120 queries ranging from easy to very complex.

Freelance Web Developer at self employed / Jun 2022 - Mar 2025.

Delivered SaaS platforms, eCommerce stores, and personal portfolios from concept to deployment. Implemented CI/CD and cloud hosting across Vercel, Heroku, and Netlify while managing projects for European clients from UX design and API integration through maintenance.

BSc in Economics, Management and Computer Science at Bocconi University / Sep 2025 - Jun 2028 (Expected).

Second-year student with a GPA of 29.86/30. Relevant coursework includes Accounting, Linear Algebra, Advanced Statistics, Computer Science, and Computer Programming. Class and course representative.

Italian Scientific High School Diploma - Applied Sciences at Liceo Majorana / Sep 2020 - Jul 2025.

Completed the Italian scientific high school curriculum with an applied-sciences specialization.

Member at BSIC Association / Jan 2026 - Present.

Contributed to research on regime-based investing, using XGBoost to select among strategies tailored to four market regimes.

Member at BAINSA Association / Oct 2025 - Present.

Coauthored a BSIC × BAINSA paper on a macro-gated SPY/GLD momentum strategy. The full backtest reported a 1.059 annualized Sharpe; a separate OVX-only test reported a 28% Sharpe improvement. Also built Metis-ChessBot, with a supervised neural evaluator and search reaching an estimated 1600–1700 Elo against Stockfish at a 125,000-node budget.

Certificate in Advanced English by Cambridge English / August 2024.

Cambridge C1 Advanced, score 199.

A Deep Understanding of Deep Learning (with Python Intro) by Udemy.

Coursework in deep-learning theory and implementation with Python.

The Web Developer Bootcamp 2022 by Udemy.

Full-stack web development course.