M.Eng. in Computer Science, Cornell University
Relevant Coursework: Advanced Artificial Intelligence, Advanced Database Systems
Bio
Download PDFEducation
May 2024
B.S. in Computer Science and Engineering, Polytechnic University of Milan
Experience
Research Assistant
Stanford University
- ●Developed autonomous AI agents to generate datasets for large-scale empirical studies.
Research Assistant
Cornell University
- ●Employed factor-based covariance decomposition and partial correlation estimators to examine correlation patterns between financing and technical debt.
- ●Designed and implemented deep learning models to classify technical debt patterns.
Founder
HESSIC Inc.
- ●Engineered multi-scale residual CNN and hybrid models for S&P 500 realized volatility forecasting, achieving up to 25.4% reductions in MSE and a 10.9% decrease in QLIKE compared to state-of-the-art HARQ models.
- ●Developed event-driven backtesting frameworks and market execution engines integrating spread dynamics and volume distribution analytics.
Data Scientist Intern
AREC-Neprix Asset Management (Milan, IT)
- ●Developed XGBoost and Random Forest ensemble models leveraging 30+ property features for distressed debt portfolio valuation.
- ●Developed a high-performance Python toolkit for portfolio analytics, leveraging memory-mapped arrays and optimized indexing for large-scale data analysis.
Data Analyst Intern
Deloitte AI & Digital Controls (Milan, IT)
- ●Engineered RPA solutions integrating rule-based orchestration, cutting transaction lead times through auto-execution workflows; implemented end-to-end BI pipelines.
Selected Projects
Task-agnostic LLM Prompt Compression
Cornell University
- ●Engineered knowledge-distillation systems leveraging teacher-student paradigm, compressing prompt tokens by an average of 43% while sustaining 87% baseline perplexity metrics.
Adversarial Machine Learning
Polytechnic University of Milan
- ●Researched and developed gradient-driven attacks as part of a 3D adversarial perturbation framework, enabling end-to-end stress-testing of model robustness.
- ●Developed a Mask R-CNN annotation system to streamline segmentation labeling.
Research Assistant
Università di Ferrara
- ●Investigated hyperparameter optimization in probabilistic logic programming.
Leadership & Service
Venture Capital Associate
Big Red Ventures, Cornell University
- ●Led technical due diligence for early-stage AI/ML ventures.
President
PoliMi Data Scientists, Polytechnic University of Milan
- ●Managed a team of 20+, raised external funds, and fostered industry collaborations.
Skills
Python, C++, SQL/NoSQL, JavaScript, LaTeX, AWS (Lambda, EC2), Git (CI/CD), Linux/Unix