In recent years, information retrieval techniques have been augmented in order to facilitate automated expert finding by searching document collections for both query topics and associated experts. Typical approaches presume that information about expert candidates exists within the document collections. I argue that such a presumption is not warranted in some enterprise settings. Instead, I turn to structured corporate data---transactions of working hours---to form document-candidate associations. Based on this idea, I design and implement a working expert finding system for a software company. I conclude that presuming the document collection to contain candidate information is not necessary in many enterprise settings because operational systems may contain structured data that is more reliable.