The scenario stores documents in a base and builds a meaning-based search: you write a plain-language query and get fitting documents.