I specialize in bringing AI-driven digitalization into real industrial environments, with a focus on improving reliability, efficiency, and sustainability in chemical and pharmaceutical manufacturing. My work translates advanced technologies like surrogate modelling, digital twins, and machine learning into practical tools for process optimization, predictive maintenance, and decision support.
At Imperial College London, I am currently developing machine learning and transfer learning methods for design space identification and process optimization in collaboration with GSK. These tools reduce experimental effort and accelerate biopharmaceutical development. Previously, I worked with Itelyum Regeneration to implement predictive maintenance strategies in a used oil refinery, increasing operational efficiency using Gaussian process regression. With SIAD Macchine Impianti, I developed a custom thermodynamic DLL in C++ to model air mixture properties, enabling more accurate air separation simulations in Aspen HYSYS and Honeywell UniSim.
I also collaborated with AVEVA to bring the digital twin experience into classrooms at Politecnico di Milano, enabling a more advanced and immersive teaching environment. Across all projects, my goal is to unlock process knowledge through real-time, data-driven solutions that support smarter industrial decision-making and improve productivity.
My approach combines hands-on implementation with a strong understanding of process systems, making digital transformation tangible for industry. Whether the goal is reducing OPEX, increasing uptime, or accelerating development timelines, I aim to deliver practical solutions that perform where it matters most, the plant floor.