The purpose of this poster is to illustrate a successful workflow for improving vendor-generated metadata for a large digital collection of archival materials by converting the metadata from the Dublin Core standard to the VRA standard using the scripting language Python.