Integration of metadata from heterogeneous sources is a major issue when connecting cultural institutions to digital library networks. Uniform access to metadata is impeded by the structural and semantic heterogeneities of the metadata and metadata schemes used in the source systems. In this paper we discuss the methodologies we applied to ingest proprietary metadata into the BRICKS digital library network and to process CIDOC CRM metadata in terms of search and retrieval, and how we strove to hide the semantic complexity from the end-user while exploiting the semantic richness of the underlying metadata.