Alessandro Flaborea

I build AI systems that understand the physical world.

As CTO and Co-Founder of Procedo, I lead the technology behind turning raw shop-floor video into structured manufacturing knowledge: extracting step-by-step procedures from operational video, grounded question-answering over a company's own documentation, and AI assistants that help industrial teams document, optimize, and train their work. It's the point where my research finally became a product people use every day.

That research is where the story starts. I recently completed a PhD at the Perception and Intelligence Lab (PINlab) at Sapienza University of Rome, working on Anomaly Detection, Action Recognition, Procedural Learning, and Hyperbolic Neural Networks. Along the way I spent time in Amsterdam as a visiting researcher, and published at top AI and computer vision venues including CVPR, ICCV, and IROS, as well as journals such as Pattern Recognition and Artificial Intelligence in Medicine.

I've always had the urge to build something of my own. Before Procedo I co-founded HomeSweatHome, an AI virtual trainer that coaches people through their workouts in real time. I was also invited to speak at Ferrari S.p.A.'s Data Science Hub on generative AI and computer vision use cases.

The common thread through all of it is curiosity: for new problems, new places, and new people. It has taken me from Udine to Rome, and from Gothenburg to Amsterdam, and it is the same thing that pulls me toward whatever I build next.