Artificial metalloenzymes (ArMs) catalyze new-to-nature reactions under mild conditions and could therefore play an important role in the transition to a sustainable, circular economy. While ArMs have been created for a variety of reactions, their activity for most biorthogonal transformations has remained modest and attempts at optimizing them by means of enzyme engineering have been case-specific and unsystematic. To realize the great potential of ArMs for biocatalysis and synthetic biology, there is a need for methods that enable the rapid discovery of highly active ArM variants for any reaction of interest. Here, we present a broadly applicable, automation-compatible ArM engineering platform. It relies on periplasmic compartmentalization of the ArM in Escherichia coli to rapidly and reliably identify improved variants based on the biotin-streptavidin technology. We assess 400 sequence-verified ArM mutants for five bio-orthogonal transformations involving different metal cofactors, reaction mechanisms and substrate-product pairs, including novel ArMs for gold-catalyzed hydroamination and hydroarylation. The achieved activity enhancements of six-to fifteen-fold highlight the potential of the proposed systematic approach to ArM engineering. We further capitalize on the sequence-activity data to suggest and validate smart strategies for future screening campaigns. This systematic, multi-reaction study has important implications for the development of highly active ArMs for novel applications in biocatalysis and synthetic biology.