Socio-technical systems are becoming increasingly complex mostly due to the unpredictability of human interactions. Furthermore, they typically work within Knowledge Intensive Environments (KIE), hence they need to deal with huge amounts of data. Coordination models are meant to cope with the increasing complexity of software systems, mostly due to the unwanted non-determinism generated by the interaction within complex systems. In this paper we describe how social actions - performed by agents interacting in a shared environment - can be exploited by a novel model for the coordination of KIE, by adopting both a nature-inspired and a cognitive/behavioural standpoint.