Content Safari is a project to research and develop methods for automating cataloguing, search, retrieval, workflow management, transcoding and rights protection of visual media held in libraries and archives. The aim is to produce a flexible, intelligent and affordable system that makes it possible for media rights owners (at any point along the chain from production to distribution) to manage and exploit content far more efficiently than is currently possible. The enabling R&D is based on automatically extracting semantic metadata from audiovisual content, using ontologies and semantic topic maps. A metadata database and workflow middleware will support the processes of profiling and retrieval, transcoding and DRM.