Rafael Felipe
11/26/2025, 1:16 PMkapa.ai
11/26/2025, 1:17 PMGROUP BY or aggregation operations at the MongoDB data source during a CDC (Change Data Capture) sync. Airbyte's CDC sync for MongoDB uses MongoDB's change streams to capture inserts, updates, and deletes, and then replicates these changes to the destination. The data is transferred as-is, without aggregation or grouping at the source level. If you need aggregated data, you would need to either create a view in MongoDB that performs the aggregation and sync from that view, or perform the aggregation as a transformation step after the data has landed in the destination system. This approach is consistent with Airbyte's ELT (Extract, Load, Transform) paradigm, where data is first extracted and loaded in its raw form, and any transformations (including aggregations) are performed after loading, typically in the destination system. If you want to be more selective about the data you sync (e.g., only sync aggregated data), you can create a view in MongoDB and configure Airbyte to sync from that view instead of the raw collection. However, Airbyte itself does not natively support pushing down GROUP BY or aggregation logic to the source during CDC syncs [Using dbt tranformation to aggregate data during a sync?](https://discuss.airbyte.io/t/3455) [MongoDB V2 Source Documentation](https://docs.airbyte.com/integrations/sources/mongodb-v2#replication-methods).Rafael Felipe
11/26/2025, 1:17 PMkapa.ai
11/26/2025, 1:17 PMkapa.ai
11/26/2025, 1:18 PM