Halim Djerroud

My research is based on the idea that an autonomous system must not only be performant, but understand what it does, why it does it, and when it should change strategy. Inspired by human cognition, I consider decision-making as a fundamentally prospective process, where anticipation, memory, and abstraction play a central role. Rather than indefinitely increasing model complexity or computing power, I explore architectures capable of reasoning, simulating, learning from experience, and interacting with humans in an explainable manner. This approach aims to design artificial systems that are more frugal, adaptive, and robust, capable of evolving in dynamic and shared environments, integrating uncertainty not as a problem to eliminate, but as a structuring element of decision-making.

Learn more about my scientific approach →