This paper describes the use of the Data Stream Processing technology on a telecommunications network to create a Real-time Mobile Advertisement System. Ads are based on the user's context, interests and profile. First we will describe the Stream Data Processing technology, its main key features, application scenarios and how it fits perfectly with the requirements posed by a proactive advertising system which seeks to provide answers to Who, When, What and How at the time of sending the ads. The When is an essential component as it provides the real-time aspect so we can take advantage of the environmental context in which the user operates. Finally we will introduce the proof of concept implemented at Ericsson Research to develop the Real-time mobile advertisement concept and its adaptation to the existing nodes in a telecommunications network. We will also describe how this implementation enables setting triggers based on location for any kind of terminal and with a sufficient positioning precision, and how Data Stream Processing handles the high data throughput involved.