This presentation addresses the question how to facilitate digital prosopography in large research infrastructures, such as the infrastructure in development by Dutch academic and cultural heritage institutions: Golden Agents: Creative Industries and the Making of the Dutch Golden Age that combines semantic web and multi-agent technologies. Such an infrastructure needs to support both researchers that are interested in individuals and those that want to focus on groups. It requires digital methods for identifying enormous quantities of names of rather anonymous people in unstructured data and link these to the more well-known names in structured data that often are used as a kind of ground truth. The latter touches upon a traditional methodological problem, i.e. the dichotomy between elite culture and the mass and the bias in data (-selection) as a result of that (Lawrence Stone, 1971). This dichotomy again comes to the foreground in digital prospopography in recent pleas for more critical approaches of individuals and social groups within the Dutch Golden Age whose wealth was partially based on slavery. With these problems in mind the general question is how can digital methods and models support prosopography in the world of big data and deal with the mixture of identified persons and of names that probably never can be linked to a person?