Universitat de Lleida (UdL), a Spanish public university, was looking to make it easier to locate among its researchers those with the expertise demanded by incoming requests from companies, institutions, and citizens in general. Information to respond to such demands was scattered among different systems, including the institutional CRIS, a DSPace publications repository, or researchers' ORCID pages. That made responding to them a very time-consuming process. To streamline the process, semantic web technologies have been used to mix all these information sources. For instance, a CERIF XML dump of CRIS data or the keywords associated with the papers in the publications repository. The output has been structured using the VIVO ontology, so it was also possible to load it into VIVO and use it as the institutional experts' guide that facilitates matching society need to university researchers' expertise. Currently, it features almost 1000 researchers, 20000 academic articles and more than 13000 research concepts. Future work focuses on curating the expertise associated with researchers, generated from those associated with their ORCID profiles or the keywords of the papers they have published. This curation is partially automated by matching these terms to concepts in Wikipedia and Wikidata, while enriching them with translations to different languages that will complement a future multilingual version of the experts portal.