Incomplete Transcript Entries immediately after transcript.v2.fileGenerated Event in Google Meet

I am observing an intermittent issue with the Google Meet API where fetching transcript entries immediately after receiving a transcriptFileGenerated event returns an incomplete list.

The Workflow

  1. Our application listens for the google.workspace.meet.transcript.v2.fileGenerated events.

  2. Upon receiving the event, we call conferenceRecords.transcripts.entries.list and iterate through all pages to retrieve the full transcript.

The Issue

Recently, we encountered a case where the API returned only 4 entries immediately after the event. However, when calling the exact same endpoint several days later, the API returned the full list of ~500 entries.

Technical Context & Debugging

  • Application Logic: We have verified our pagination logic and error handling. No possible causes were found.

  • After that event, we have begun logging the state from the conferenceRecords.transcripts API. Out of ~10,000 processed events, the state has consistently been FILE_GENERATED.

Questions

  1. Does receiving the google.workspace.meet.transcript.v2.fileGenerated event guarantee that all transcript entries are available via the entries.list method? The documentation states that it indicates the transcript file is generated, but I couldn’t tell whether it also means the API is ready to be called.

  2. Are there known edge cases where the file generation event is fired before the entries sub-resource is fully populated?

Any insights or recommendations on how to handle this more robustly would be greatly appreciated!

1 Like