@ValleyMountWalker It sounds like there may be a few different issues affecting the call recordings and transcriptions. Here are some potential causes: 1. **Truncated Recordings**: If the recording starts late or misses sections, it could be due to the voice activity detection (VAD) settings. VAD is designed to detect when someone starts speaking, but if it's not calibrated correctly, it might miss the beginnings of sentences or entire sections if the speech is too quiet. 2. **Overlapping Speech**: If the agent and customer appear to be speaking at the same time in the recording but were actually speaking at separate times, this could be a result of audio channel mixing issues. Ideally, the agent and customer should be recorded on separate audio channels to avoid this. 3. **Transcription Inaccuracies**: If the transcription doesn't reflect the full call flow, it could be due to background noise, poor audio quality, or the transcription model struggling with accents, speech patterns, or technical jargon. To address these issues, you may want to check the following: - Ensure that the VAD settings are correctly configured to capture speech as soon as it starts. - Verify that the recording setup is correctly separating the audio channels for the agent and customer. - Consider using noise reduction techniques or improving the acoustic environment to enhance audio quality. - Review and adjust the transcription model settings, or provide additional training data to improve its accuracy with your specific use cases. If you have access to the system settings or can provide more details, I can offer more targeted advice. Otherwise, it may be best to reach out to the support team for a deeper investigation into these issues.

You can continue this conversation by mentioning me in the message.

can you let me know a good time for you to come on a call?