I am a postdoc researcher in Computer Science with a background in Mathematics.  I mainly work on probabilistic programming (analysis, inference, synthesis, differentiation, etc.) and more in general on the approximation of probabilistic systems.

Short Bio 📕

Since September 2025 I am a Postdoc Researcher at TU Wien in the TrustCPS group led by prof. Ezio Bartocci, where I am the holder of a MSCA Post-doctoral fellowship for the project EMBEr.

Previously, I was a Postdoc Researcher at the University of Trieste, where I was also holder of a scholarship for the project "Program synthesis and automatic differentiation for probabilistic programs" (scientific supervisor: prof. Luca Bortolussi).

I have received my Ph.D. in Computer Science and Systems Engineering at IMT School for Advanced Studies Lucca, with a thesis titled "Efficient and accurate analysis of of two classes of generative models", written under the supervision of prof. Mirco Tribastone (IMT Lucca) and prof. Luca Bortolussi (University of Trieste). The thesis was recognized as Best Italian Ph.D. Thesis in Theoretical Computer Science by the Italian Chapter of EATCS. 

During my Ph.D. I visited for three months the MOVES group at RWTH Aachen, under the supervision of prof. Joost-Pieter Katoen (RWTH Aachen).

Before my Ph.D., I got my MSc. in Mathematics at University of Trieste, where I was the holder of a scholarship from SISSA. I got my BSc. in Mathematics at University of Catania where I was a student at Scuola Superiore di Catania.

Research Interests 🔎

I am interested in studying efficient ways of performing inference on probabilistic programs involving continuous distributions. During my Ph.D. I developed a method to perform inference based on Gaussian Mixtures and moment-matching. The SOGA tool (2023), implements a particular instance of this method. The DeGAS tool (2026) enhances this method allowing for optimizable parameters and sample-free gradient-based optimization of probabilistic programs. 

I am actively working in applying these tools to various topics such as the synthesis of probabilistic programs, quantitative information flow and analysis of cyber-physical systems. 

Previously during my Ph.D. I worked on a new approximation method for Markov Population Processes, that refines classical mean-field techniques even for systems exhibiting non-differentiable drifts and oscillatory behaviours. The reference paper was published in SIGMETRICS in 2021.

See my list of publications here.

If you think we can collaborate, feel free to contact me! 😊

(Co)-Supervisions 🎓

• Ph.D. Students: Romina Doz (with Laura Nenzi, University of Trieste)

• B.Sc. Students: Oguzhan Mertturk (with Ezio Bartocci, TU Wien)

I am available to (co-)supervise the following two thesis at TU Wien:

• Privacy Metrics Computation for Prob-Solvable Loops  (Master Thesis)

•  Comparing Grammar Based Evolution and Monte-Carlo Markov Chain for the Synthesis of Probabilistic Programs (Master Thesis)

Academic Service 🏫

• 2026: VeriProp (co-chair), ATVA (PC), QEST (PC), WADT (PC), OOPSLA (AEC)

• 2024: QEST+FORMATS (AEC)

Teaching and Tutoring 🧑‍🏫

In 2026 I have thought the mini-course "Python Programming" at the Winter School STUDIOLO - Digital Humanities Lab.

In the academic year 2024/2025 I have been tutor for the course "Algoritmi and strutture dati" from the Bs.C. in Ingegneria Elettronica and Informatica at University of Trieste, taught by prof. Luca Bortolussi.

Other interests 👾

Besides research I am keen on science communication!

In 2020 I was one of the finalist of the FameLab Italy competition. Moreover, I was a speaker coach for TEDxCataniaWomen and TEDxCataniaSalon. I have written for different online blogs and magazines (see, for example OpenIMT, Treccani, Antropia, Meer) and I was part of the project La Sagra della Scienza.