looking for advice on a specific gap in our bulk export pipeline for legal discovery requests. we are running a predictive routing strategy in genesys cloud eu-west, utilizing multiple skill-based queues with distinct service levels. the issue arises when triggering a bulk export job via the recording api for interactions that were initially routed through these predictive queues.
while the audio files and basic transcript metadata are successfully pushed to our s3 bucket, the specific queue assignment data and the predictive routing decision logs are not present in the exported csv files. for standard manual routing, the queue_id and routing_reason fields are populated correctly. however, for interactions handled by predictive routing, these fields appear null or contain generic placeholders like “auto-routed”. this creates a significant problem for our chain of custody documentation, as we need to prove which specific queue configuration was active at the time of the interaction for compliance audits.
we have verified that the oauth token has the necessary permissions for recording:read and routing:read. the export job completes with a status of 200 ok, and no errors are logged in the job history. the s3 integration is functioning correctly for other data types. we are using the standard bulk export endpoint /api/v2/recording/bulkexportjobs. has anyone encountered similar issues with predictive routing metadata not being included in bulk exports? is there a specific flag or additional api call required to include the routing decision history in the export payload? we need to ensure that the metadata aligns with the interaction timeline for legal hold purposes. any insights on whether this is a known limitation of the current api version or a configuration issue within the routing settings would be greatly appreciated.