Social media users have shifted their communication medium to mobile apps, especially when important decisions need to be made at anytime, anywhere. Often at times, virtual groups are created over mobile apps to brainstorm on solutions to the problems and decide on a right solution. The process of making decisions over the mobile apps group could be analysed by identifying the pattern of group interaction. Among the decision-making techniques identified for analysis are the consult and consensus, which could further contribute in identifying the leadership type, observing member and non-observing member that exist in a virtual group. This paper proposes a deployment of personal knowledge management processes on the mobile apps group interaction, for a case of decision-making process. The outcome includes an agent-mediated knowledge management model that introduces Monitoring Agents, with roles to get, understand and share knowledge on interaction patterns, and further identify the leadership type, observing member and non-observing member. Future work is seen as beneficial to agent-mediated decision support system, in which the Monitoring Agents could assist in simulating the right decision-making technique to use for solving different kinds of problems.