CV_Antonio_Pelusi

Summary

Researcher and software developer focused on the intersection of advanced generative AI, structured data synthesis, and privacy-preserving Machine Learning systems.

Investigates the structural limitations of generative architectures on structured data, combining rigorous evaluation of statistical fidelity and memorization risks with practical system development.

Experience

Random Power (Startup)

| May 2024 - Nov. 2025 Full Stack Developer | Remote

Developed a comprehensive Django-based web client (backend and frontend) for Random Power's 64x Multi-Generator Board quantum hardware platform, enabling secure multi-user access with authentication and session management to true quantum randomness generation.
Built Palo Alto QRNG API-compliant Entropy-as-a-Service platform supporting 1Gbit/s quantum random bit distribution with automated quality assurance and real-time compliance monitoring.
Collaborated in the design and development of scheduler and TEE components for secure random bit distribution.
Designed and maintained communication interfaces between hardware and software components, ensuring reliable and secure data exchange.

A. Paradisi High School

| Sep. 2024 - Jul. 2025 Computer Science High School Teacher | Vignola, Italy

Introduced 75+ high school students to programming, algorithms, and fundamental computer concepts using practical exercises and projects.

University of Modena and Reggio Emilia

| Sep. 2024 - Dec. 2024 Computer Science University Tutor | Modena, Italy

Conducted laboratory sessions, guiding 50+ undergraduate students in programming, algorithms, and data structures.
Provided hands-on support during labs, assisted with debugging, and facilitated practical understanding of theoretical concepts.

Liferay Inc.

| Mar. 2021 - Jul. 2021 Junior Full Stack Developer (Internship) | Remote

Integrated Stripe payment functionality into the open-source Liferay Portal.

Education

Ph.D. in Computer Science and Mathematics of Computation

| Nov. 2024 - Present University of Insubria

Researched privacy-preserving synthetic data generation with small LLMs, investigating LoRA fine-tuning and Differential Privacy techniques for tabular data synthesis.
Researched Differential Privacy frameworks analyzing how quantum vs. pseudo-random number generation impacts privacy mechanism effectiveness.

Master's Degree in Computer Science

| Sep. 2022 - Apr. 2024 University of Modena and Reggio Emilia | 110/110

Big Data & Machine Learning: Focus on predictive modeling, large-scale data analysis, and advanced machine learning methods.
Cybersecurity & Cryptography: Exploration of theoretical and applied aspects of data security and cryptographic protocols.

Post Graduate Degree CBI.Attract

| Feb. 2023 - Jun. 2023 University of Bologna, CERN, AGH University | 30/30 with Honors

Participated in research collaborations at CERN and AGH University, funded by the EU Attract Project, contributing to joint investigations on emerging technologies.
Explored technological opportunities through Design Thinking-based prototyping and testing.

Bachelor's Degree in Computer Science

| Sep. 2018 - Apr. 2022 University of Modena and Reggio Emilia

Computer Science Foundations: Algorithms, data structures, programming languages, computer architecture, and operating systems.
Mathematical Foundations: Discrete mathematics, calculus, statistics, and linear algebra.

Publications

Privacy Preservation in Data Sharing: A Study on Differential Privacy

| Apr. 2024 Master's thesis

An introductory and analytical study of differential privacy as a technique for preserving data confidentiality.

Projects

chesSLM

| Jun. 2026 Chess Small Language Model: a Decoder-Only Transformer Built From Scratch to Play Chess via Move-Level Language Modeling

Designed and implemented from scratch a ~100M-parameter decoder-only Transformer (custom attention, training loop, and tokenization scheme), trained via a two-stage pretraining/ finetuning curriculum and deployed as a UCI-compliant chess engine.

LLM-synthetic-data-generation

| May 2026 LLM-based Synthetic Tabular Data Generation Pipeline

Designed and implemented a config-driven pipeline to generate and evaluate synthetic tabular data using large language models with LoRA fine-tuning and differential privacy.

epsiloc-lib

| Oct. 2025 Local Differential Privacy library

Designed and implemented a library for applying Local Differential Privacy mechanisms to local databases.

Awards

Winner of Green & Digital Transformation Hackathon

| Sep. 2023 University of Modena and Reggio Emilia | Mantua, Italy

Winner of a hackathon with a machine learning-based predictive model for water consumption optimization.

Skills

Backend: C, C++, Java, Kotlin, Python - Django, Android SDK - SQL (PostgreSQL, SQLite), NoSQL (MongoDB, Neo4j)

Frontend: HTML, CSS, JavaScript, ImGui, JavaFX

Libraries: Pytorch, SKLearn, NLTK, LangChain, OpenDP, diffprivlib, OpenMP, CUDA, Pandas, Numpy, MatPlotLib, Seaborn

Tools: Linux, Git, Docker, Make, Gradle, Maven

Languages: Italian (native), English (B2)

Antonio Pelusi

Ph.D. student in Computer Science and Mathematics of Computation