As a researcher at CIRAD since 2001, my work focuses on participatory modelling. This involves establishing foresight approaches supported by models designed and used collectively. The aim is to design (1) tools and methods to facilitate the collaborative design of multi-agent models and (2) tools to foster interactions with a simulation to influence simulated dynamics. As such, the Cormas platform for which I am responsible, has the triple role as i) a research tool to assist the participatory design of models to improve their uses through interactive simulation, ii) an operational tool on the field and in projects and, iii) as a learning medium in training sessions, such as summer schools and university courses. I coordinate the development of Cormas, which requires new IT developments to make it operational in various contexts (land-use, hydrology, aquifer and livestock management, etc.). From a technical perspective, the reason for including UML diagram editors is to facilitate and accelerate model design and implementation. The simulation component addresses the dynamics of social and ecological systems, in particular by considering the development of new tools for interactive simulations (distributed simulations, information asymmetry, agent manipulation, backward simulation, sensitivity analysis, etc.). Beyond the tools, my experience in the field contributes to reflections on how to conduct modelling workshops with stakeholders, and on the role of modelling as a key component for interdisciplinarity. In contrast to conventional approaches, my research seeks to understand the advantages and limitations of simulator-based foresight studies. In doing so, I seek to contribute to ethical issues on how to engage stakeholders, as well as the social and political responsibility of the facilitator. Within the Green research unit of CIRAD, I provide input on issues such as the analysing tensions between environmental conservation and development, as well as the role of collective action. Rather than proposing ‘turn-key’ solutions for development, the ComMod approach, in which I am involved, can play a major role in helping a group of actors to manage their resources. By reflecting on the context of his intervention and trying to understand the power games, the modeller can, in certain circumstances, encourage the emergence of a shared vision between the stakeholders on their own socio-ecosystem. The model thus becomes a mediation tool and its long-term projection can change perceptions on individual and collective practices. This role of the approach as catalyst can lead the stakeholders to establish their own collective rules for the viable management of their commons.