I am currently an Associate Professor at the Department of Physics and Astronomy of the University of Bologna and a member of the ATLAS Collaboration. After obtaining my PhD in Physics from the University of Bologna in 2014 and spending four years at CERN in Geneva, from 2012 to 2016, I returned to Bologna, where I have pursued a research line combining experimental high-energy physics, the development of advanced reconstruction and data-analysis algorithms, machine learning, HPC and, more recently, quantum computing applied to particle physics.
Within ATLAS, I work on machine learning and HPC for the development of advanced deep learning models for the fast simulation of the ATLAS calorimeter, in particular in the context of FastCaloSim and FastCaloGANtainer, a framework of which I am a co-author. This activity aims to make fast simulation more accurate, scalable and efficient, also in view of the computational challenges of the HL-LHC. I have contributed to the porting and optimization of GAN training on high-performance computing infrastructures, including Leonardo-CINECA and CNAF-HPC, exploiting multi-GPU A100 nodes and significantly reducing training times compared to traditional configurations.
In parallel, I work on the development of neural networks and advanced models for jet tagging and anomaly detection, applied to several physics analyses. These activities range from precision measurements of the top-quark cross section and of the properties of the Higgs boson to searches for new physics involving type-I/III heavy neutrinos, vector-like leptons and dark matter candidates such as dark jets. In this context, I also explore graph-based architectures and diffusion models, with the goal of improving the identification of complex signals and rare topologies in the high-luminosity regime.
My activity in ATLAS also includes the development of reconstruction algorithms for boosted jet topologies and for hadronic decays characterized by the presence of muons. This line of work, initiated during my time at CERN, is part of the broader effort of the Collaboration to optimize the reconstruction and selection of complex final states, relevant both for precision measurements and for searches beyond the Standard Model.
Since 2025, I have coordinated the Bologna unit of the QUART&T project, Quantum Architectures for Theory & Technology, dedicated to quantum computing. Within the project, I work on the development of demonstrators and simulators of quantum circuits based on superconducting qubits in resonant cavities, with all-to-all connectivity architectures, and on qudits realized through Josephson junctions. The activity also includes the design and integration of RF-SoC read-out electronics, with the goal of developing experimental platforms and control tools for superconducting quantum devices.
Alongside hardware and simulation development, I work on the design of quantum algorithms and, in particular, quantum machine learning (QML) applied to particle physics. In this area, I have contributed to QUnfold, an open-source tool that reformulates the statistical unfolding problem in terms of QUBO, making it solvable on quantum annealing hardware. This activity lies at the intersection of statistical inference, quantum optimization and data analysis in HEP, with the aim of exploring the potential of quantum technologies for computationally complex problems in experimental physics.
Finally, I participate in the FOOT Collaboration, which measures fragmentation cross sections of light nuclei for applications in hadron therapy and space radiation protection. In FOOT, I have contributed to nuclear-fragmentation data analyses and to the implementation of the Kalman-filter-based tracking system, developing tools for track reconstruction and the analysis of experimental events.
