In recent years there is a rapid growth of unstructured text content and multimedia documents, which includes audios, videos and images in web as well as in the enterprise and these large volumes of data is not very useful without effective methods for content analysis and retrieval. In this presentation we present Sensefy – a cross media information retrieval system for enterprise, which uses higher-level semantic concepts and entities for retrieving documents of different media types. Here we describe the process of automated semantic information extraction from multimedia documents, which is central in developing concept and entity driven search across different media types. Media In Context(MICO) platform is described as a platform for cross media analysis.