Courses
Introduction to Quantum Computing + AI, with a focus on applications
This course — which I currently teach at UAI, as a diploma program — is split into four modules: an overview of applications in quantum technologies, foundations of quantum physics and computing, quantum programming and software, and technology-watch tools. The goal is for anyone interested in quantum technologies to come away with a basic sense of “what quantum tech is and what it’s for”, along with the fundamental tools for programming it. To get there, we hold on to a very current topic, “quantum deep learning”, and ask how you’d define, program, and train a neural network with a quantum circuit if you wanted to apply it to a specific problem. Along the way, we look at both the possibilities and the practical drawbacks of actually doing this.
Theoretical and practical introduction to reinforcement learning
This course runs for 16 classes of 2 hours each and aims to teach the fundamentals of Reinforcement Learning, from bandit theory to PPO, passing through Q-learning, REINFORCE, and the fundamentals of Markov Decision Processes. Throughout, we emphasize algorithmic applications to concrete problems, and the course is accessible to anyone interested in machine learning, with no particularly strong math background required. It wraps up with a review of modern applications of deep reinforcement learning, including RLHF, RLVR, breakthroughs like Jev, and the relationship with world models.
The impact of quantum technologies on industry
This course runs over three two-hour sessions — put together with Practia Global — and offers a self-contained, critical, general view of quantum technology applications (computing, sensing, simulation, communications) in industry. It focuses on different business models and their potential impact on emerging economies.
AI for scientific research
This course provides the fundamental tools and knowledge to bring AI into scientific research. In the first part (foundational concepts), we start from small, tabular models to tell apart different primitives (supervised, unsupervised, reinforcement learning), then move on to deep learning and the most popular architectures (MLPs, CNNs, RNNs, transformers, U-nets, diffusion models), up to programming an LLM and, finally, an agent. In the second part, on tools, we bring in technology-watch techniques so students can keep up with new developments; in the 2026 edition of the course we cover the fundamentals of agentic AI, both claude-code and Hermes, as well as prompt & context engineering and multi-agent coordination and orchestration schemes for repetitive tasks.
AI for everyday life
This training — put together with LIAA and Anthropic through the “Anthropic ambassadors” program — provides the fundamental tools for bringing artificial intelligence into different processes, both personal and business. We look at how to design a digital assistant thoughtfully (after first identifying the actual need for one), as well as how to build automations in N8N, API integrations, and models for evaluating the return on automation.