This Chapter examines the artworks of several contemporaries of William Billingsley who painted at the Derby China Works and elsewhere during the late Eighteenth and early Nineteenth Centuries, with the aim of establishing robust reference databases for the application of AI-based machine learningMachine learning authentication. The chapter opens by contextualising the stylistic proliferation of Billingsley’s characteristic floral compositions after his departure from Derby, noting the continuation of his popular patterns—most notably the Prince of Wales dessert service—under successive factory managements and with different pattern numbers, a practice that has generated considerable misattribution in subsequent connoisseurshipConnoisseurship. The critical importance of selecting only unequivocal “autograph” exemplars for the AI standard database is then addressed in detail, and the potential pitfalls involved in this selection process are systematically identified: these include workshop collaborations, the continuation of popular service patterns by unnamed copyists, the decoration of pieces from manufactories with which an artist had no formal connection, and the presence of fakes and forgeriesForgery. Autograph exemplars of four of Billingsley’s principal contemporaries are then presented and discussed: Edward WithersWithers, Edward, his early mentor at Derby; William “Quaker” PeggPegg, William \“Quaker\”, celebrated for his bold botanical renditions; Moses WebsterWebster, Moses, whose pink rose groups most closely approach the Billingsley idiom; and Leonard LeadLead, Leonard, a highly accomplished Derby decorator working in a comparable style. The chapter demonstrates that rigorous exemplar selection, informed by both documentary evidence and connoisseurial consensus, is an essential prerequisite for training reliable AI classification algorithms, and that the stylistic proximity of these artists to Billingsley makes their inclusion in the comparative database particularly valuable for the disambiguation of contested attributions.