Watson is a gateway to the Semantic Web: it collects, analyzes and gives access to ontologies and semantic data available online. Its objective is to support the development of next generation Semantic Web applications that dynamically select, combine and exploit the knowledge published on the Semantic Web. We present the design of Watson, which have been guided by the requirements of these applications and by lessons learnt from previous systems. In particular, the three main activities handled by Watson are detailed: crawling the Semantic Web, analyzing and indexing semantic data and ontologies, and providing efficient access to this data to web users and applications. In addition, some novel Semantic Web applications built thanks to Watson are briefly presented. KEYWORDS Semantic Web, gateway, search engine, ontologies, RDF, query.