Semantic layers centralize metric definitions and deliver consistently calculated metrics to data consumers, but without proper tooling, entropy can prevail inside the layer's data model. Cube Cloud aims to address this by providing a data model editor, version control support, and development mode with code branches. However, source code is not the only representation of the data model, and humans may have different needs than machines in understanding and working with it. To cater to both, Cube Cloud introduces Data Graph, an entity relationship diagram that uses crow's-foot notation to mark join relationships between cubes, allowing users to visually explore and reason about the data model in a human-friendly way.