From MongoDB

DocumentForge will feel familiar: JSON documents grouped into collections. The biggest change is querying — you write SQL (with nested-path and array access) instead of the aggregation pipeline, and you declare indexes on the paths you filter by.