Rationally designed protein domain mimics are based on secondary and tertiary structure epitopes observed at protien-protein interfaces. Successful development relies on interplay between good quality crystal and computational data, with iterative synthesis of scaffolds displaying key residues involved in binding. These direct mimics are limited in their binding affinity by the functional transience and low affinity of the native complex and require significant optimisation. We seek to translate an optimised tertiary mimic scaffold, a crosslinked helical dimer, to a phage screening platform that would allow access to diverse libraries of conformationally defined interfaces. The libraries could then be refined against a target of interest and amplified for further rounds of selection, deviating from the need to define the target epitope by structural analysis. With our developed library, we screened against VEGF, identifying 8 unique sequences which have been shown to bind by ELISA, and in vitro binding studies using fluorescence polarisation are underway.