This paper aims to present the development of Single Agents for the energy management of electric energy sources including PVs in buildings from the one hand and acheiving the implementation of Demand Side Management strategies on the other. Actions of electric utilities by TSOs/DSOs could influence the amount or timing of customers' energy demand, in order to utilize scarce electric supply resources more efficiently. Using Smart Metering, consumers receive signals that reflect the actual price of electricity, which fluctuates in response to supply and demand. A Single Agent Thermal Model-Based Predictive controller is developed implementing super cooling/heating strategies, through which a building or a building part is pre-cooled/pre-heated during low peak periods achieving peak shaving in an acceptable way by the users regarding the thermal comfort issue. A pilot application of Single Agents is implemented in a building in Greece, and the results are analysed and presented.