Abstract : As buildings contribute significantly towards global energy consumption, it is essential that the occupants receive the best comfort without utilising further energy. This work treats building, environment and the occupants as a system, which presents the context and the occupants also provide their comfort criteria to a black box for yielding the a schedule of actions (opening/closing of doors/windows) for optimal comfort. The physical state of an office, situated in France, is recorded over a span of 100 days. This data is utilised by a physical model of the building to simulate the indoor ambience based on random sets of user actions from which an optimal schedule is obtained, representing equally best trade-off among minimal thermal and CO2 based air quality dissatisfaction. Results indicate that adopting the proposed schedule of user actions can efficiently enhance the occupants comfort.