The monitoring of print and online media has been taking place more or less automatically for a long time. Since an ever greater proportion of new content is produced first or even exclusively for audiovisual (AV) media, the importance of monitoring these AV media has soared. Since the monitoring of AV media can hardly be done manually, the monitoring process has to be automated as much as possible, especially the selection of relevant contributions.The goal of this innovation project is therefore to research, design and develop an innovative and efficient AV media monitoring pipeline with a focus on Swiss AV media whose content is a combination of Swiss dialects, German, French, Italian, Rhaeto-Romanic, English and other languages. In this project, the first part of the pipeline will be automated as much as possible by a novel method of STD (Spoken-Term-Detection) for the identification of relevant media content as well as an optimized Speech-To-Text system for spoken spontaneous speech as it occurs in Swiss AV media.