The increase in demand for enhanced building performance is becoming a challenging task to fulfill especially during buildings operational phase. This situation creates the need for the development of new process controls for buildings operations to achieve the goal of a completely intelligent building which is able to respond dynamically to internal and external perturbations in order to foster comfort conditions for the building occupants. Unpredictable user-behavior, changing weather conditions, mismatch between consumption and production of resources, operation of active and passive climate-control systems are factors that affect buildings behavior hence requiring the use of intelligent control decisions in realtime. These intelligent control decisions directly affect the buildings energy efficiency and occupant thermal comfort The road towards intelligent control decisions in real time for intelligent buildings operation requires groundbreaking innovations and progress beyond the current state-of-the-art in various aspect of building construction. In achieving the desired goals, the application of AI techniques promises to be very useful, especially multi agent systems which has shown enormous potential for improving the operational control of buildings. Up until now, there has not been any real application of intelligent agents in building management systems, however field tests with multi agent systems on top of current building management systems has shown to be a promising approach for combined cooperation between building management systems and the Grid. A further development is to implement MAS structures for comfort management in combination with energy optimization as an addition to existing building management functionality. In this paper, an ontology for open multi agent platform in combination with a Building Energy Management System is presented.