Strategy, market analysis, and consulting projects from academic and professional engagements with real companies.
Designed a 3-pillar AI credit scoring engine for UniCredit to bridge the information gap for 4.4 million Italian SMEs — combining traditional financials, digital footprint, and growth potential with dynamic weights and explainable AI to reduce Cost of Risk by 10–20%.
Designed a data-driven Spain expansion strategy for ProprioGO — a French real estate platform — including city prioritization, €500K budget allocation, and a CPA forecast of €5.81 across 5 target cities.
Built a top-down, metrics-driven scoring model to prioritize Moncler's next store openings across Retail, Grenoble (ski resort), and Pop-up formats — recommending Dublin and Luxembourg City for retail, Cervinia for ski, and Livigno for the 2026 Winter Olympics pop-up.
Transformed raw, unreliable substation energy data into a 100% complete, model-ready dataset — removing 1,493 duplicates, reconstructing 2,620 missing intervals, and engineering 22+ predictive features for high-precision anomaly detection.
