I’m Hajar Goddi, and my PhD sits at the crossroads between human interaction analysis and artificial intelligence, within the Critical Metals in Orogens (CMiO) project. While some of my colleagues study the rocks and resources of the Massif Central, my fieldwork takes place in meeting rooms,  where I explore how groups communicate, cooperate, and negotiate around complex scientific and societal issues.

Understanding these dynamics is essential: the transition to sustainable resource management doesn’t rely only on technology or geology, but also on our ability to work together and to build consensus.

Using tools from AI and deep learning, I aim to automatically model group behaviors : by learning patterns of interaction, synchrony, and engagement directly from audio and video recordings. Automating this analysis helps us move beyond the limits of manual annotation and subjectivity, paving the way for a deeper, data-driven understanding of collaboration and dialogue  a key ingredient for the success of interdisciplinary projects like CMiO.