I would like to be able to configure backup snapshots of a single database or collection on a schedule or continuously. This feature would help projects that need snapshots but only on a subset of data. Cluster-level, topology dependent snapshots may not be appropriate or make sense financially for all projects.