Context:
Need some help troubleshooting a discrepancy in our bulk recording exports. We are operating in the Europe/London timezone and recently initiated a legal discovery request for WhatsApp interactions. The goal is to retrieve all voice recordings and associated transcripts from a specific date range for a compliance audit. The export job was triggered via the /api/v2/recordings/bulkexport endpoint using a POST request. The request body included filters for recording_type: "digital" and a specific date range. The job ID was returned successfully, and the status eventually moved to COMPLETED.
However, upon downloading the CSV manifest and the corresponding audio files from the S3 bucket, a significant mismatch in metadata was observed. The start_time and end_time fields in the CSV manifest are approximately 15 minutes earlier than the actual timestamp embedded in the WhatsApp message metadata retrieved via the Conversations API. Furthermore, some recordings linked to the same conversation ID appear to have different durations in the manifest versus the actual file properties. This inconsistency poses a serious risk to the chain of custody for the legal hold. The environment is using the latest available API version, and the S3 integration is configured with standard permissions. No errors were logged in the Architect flow or the export job history, other than a warning about “potential timezone offset adjustments” which was ignored initially.
Question:
Can anyone clarify if there is a known issue with timezone handling for digital channel recordings in bulk exports? Specifically, does the start_time in the export manifest reflect the server-side processing time rather than the actual client-side message timestamp? We require the exact client-side timestamps for legal validity. Is there a way to force the export to include the original conversation metadata timestamps instead of the normalized server times? Any insights on how to reconcile this discrepancy without re-running the entire export job would be appreciated. We need to ensure the audit trail remains intact and accurate for the discovery process.