In a lot of countries many of existing buildings are not equipped with a mechanical supply-exhaust ventilation. Even in the case of renovation, it is not technically possible or economically feasible to design and install such a ventilation system. Therefore, a solution with natural ventilation is often the only solution. As the natural ventilation is not predictable, dependent on external forces, and requires involvement from personnel using them by manual window opening it would be beneficial if this process would be automated and based on the actual needs of the premise. This study presents information of a developed automatic window control system which is based on real-time room measurements. Information about the inside and outside air temperature, CO2 levels, as well as human location in the room was gathered through the use of sensors and a camera. The obtained image is led through pre-trained AI which recognizes the persons and software which converts the detected persons to a point in the X-Y coordinate plane. Afterward, this data is going to our developed algorithms, and the system autonomously regulates window operations to minimize disruptions to occupants' thermal comfort. The gathered data regarding the system's performance is systematically analyzed to see if the system is working as intended and can provide adequate indoor air quality at the same time not compromising thermal comfort. The results show that the developed system effectively identifies individuals within a room, determines their relative positions in relation to operable windows, and continuously monitors temperature, relative humidity, and CO2 concentration. The IAQ measurements were coupled with occupant surveys regarding the noise and disturbance caused by system operation as well as IAQ and thermal comfort. The results showed that the opening of the windows did not cause any noise issues at the same time some persons noted that there was a cold draught. The system was tested when the outside air temperature was around +5°(+41°F).