Anyone free to help troubleshoot this persistent issue with our automated recording exports for legal discovery. We are using the Genesys Cloud Recording API to trigger bulk export jobs that include both the audio file and the AI-generated transcript metadata. The setup involves a direct S3 integration via the POST /api/v2/recording/bulkexport endpoint.
The job initiates correctly with a 202 Accepted status. However, when the export type includes transcript, the job status remains in PROCESSING for over 4 hours before eventually failing. The failure reason provided in the job details is vague, citing a generic METADATA_EXTRACTION_ERROR. This is problematic for our chain of custody requirements, as we need the transcript timestamps to align precisely with the audio for audit trails.
Here is the error payload from the job summary:
{ "id": "exp_99887766", "status": "FAILED", "reason": "METADATA_EXTRACTION_ERROR", "details": "Unable to serialize transcript entities for S3 upload." }
We have verified that the S3 bucket permissions are correct and that standard audio-only exports succeed without issue. The environment is running the latest release. Has anyone encountered serialization issues with transcript entities in bulk exports recently? We are specifically looking for workarounds or known limitations regarding AI transcript metadata in the export pipeline.