This presentation shows the application of semantic technologies to company internal data sets. We are calling our approach “semantic roundtripping”. There are four specific aspects of this approach. First, we focus on semantic processing of XML data sets, using format preserving enrichment. Second, we are careful not to break existing XML based workflows. Previous industry projects that tried to bridge the worlds of semantic technologies and XML have shown that the overhead of adapting an existing workflow can be a killer argument against deploying semantic technologies in a real-world industry scenario. Third, one key motivation for our work is the integration if data across companies. With common data warehousing approaches, this is hard to achieve. Hence, we are using Linked Data as a means for secondary information modelling and data integration. Finally, we are targeting XML based formats and workflows that are at the core of existing business processes in the participating companies. These companies are dealing on an everyday basis with highly specific workflows and data formats. Only in this way it is possible to achieve roundtripping during the application of semantic of semantic technologies.