The diploma thesis describes an intelligent ambient system, the center of which is an algorithmic generator of ambient music. It is created based on a combination of established algorithms for generating music. Through the time of its operation it generates a continuous flow of ambient music and at the same time enables control over the properties of the generated music. The properties of the music can therefore adapt over time to the properties of the space in which the installation is located, for example to the number of visitors, time of day, loudness of the space, etc. The mapping is provided by a feedback loop, through which listeners give an opinion on the appropriateness of the music. The basic set of feedback data was obtained with the help of an online survey. Based on the survey results, we distinguished certain connections between the parameters of the generator and the measurable parameters of public spaces. In the future, the goal is to set up the system in a public space to provide a continuous original music background for visitors and to capture new feedback.