Offensive security is where I come from: penetration testing, source code review, secure design assessment, infrastructure, and engagements led end to end. It is the ground the AI work is built on.
Most of my attention now goes to AI security: red teaming GenAI and LLM-integrated applications for prompt injection, indirect context manipulation, and model-level abuse. The same interest runs the other way too, using AI to make offensive work faster, which in practice means agentic orchestration and internal tooling.
Client work stays under NDA, so what I publish comes from the research I run alongside it. That is where the passkeys series and the DEF CON 34 talk came from.
Academic background in Computer Engineering and Security: BSc at UniRoma3, then PoliMi for the MSc, to specialize in cyber security.
Jazz guitar since I was nine, six of those years at Saint Louis College of Music. These days a hobby.
Here you can find an extract of my experience and education. The full CV is available upon request, please contact me first.
Selected Work
Record of Employment
Certifications & Licenses
Main Education
Additional Education
Volunteering
