Online retailing has been experiencing explosive growth for years and is dramatically reshaping the way people shop. Given the lack of personal interactions fashion retailers have to establish compelling service and information offerings to sustain this growth trajectory. A recent manifestation of this is the emergence of shopping curation as a service. For this purpose, experts manually craft individual outfits based on customer information from questionnaires. For the retailers as well as for the customers, this process entails severe weaknesses, particularly with regard to immediateness, scalability, and perceived financial risks. To overcome these limitations, we present an artificial fashion curation system for individual outfit recommendations that leverages deep learning techniques and unstructured data from social media and fashion blogs. Here, we lay out the artifact design and provide a comprehensive evaluation strategy to assess the system's utility.