Abstract. During the last years a series of repositories containing learning contents for architecture have been created. With all the repositories financed, designed, implemented and maintained independently, valuable and expensive information contained in the repositories lies — scattered over several European countries — in several thousand learning objects. As every repository holds its own set of learning objects, uses different metadata schemas, and is not being connected to others, the contents are unnecessarily hard to find and cannot be used to the full extent possible. This is partly due to legal and to technical difficulties. In MACE, we aim to overcome these issues and create an infrastructure to enable the discovery and identification of learning objects all across different repositories in a uniform way. Key words: e-learning, content enrichment, content repurposing, metadata, learning object discovery, information retrieval, architecture, MACE