Quickly navigate transcripts and audio - Amazon Connect

Quickly navigate transcripts and audio

Supervisors are often required to review many agents calls for quality assurance purposes. The turn-by-turn transcript and sentiment data helps you quickly identify and navigate to the portion of the recording that is of interest to you.

This topic explains how you can quickly navigate transcripts and audio to find areas that need your attention.

Turn on autoscroll to synchronize the transcript and audio

Autoscroll enables you to jump around the audio or transcript, and the two always stay in sync. For example:

  • When you listen to a conversation, the transcript moves along with it, showing you sentiment emojis and detected issues.

  • You can scroll through the transcript, and choose the timestamp for the turn to listen to that specific point in the recording.

Because the audio and transcript are aligned, the transcript can help you understand what the agent and customer are saying. This is especially useful when:

  • The audio is bad, maybe due to a connection issue. The transcript can help you understand what's being said.

  • There's a dialect or language variant. Our models are trained on different accents so the transcript can help you understand what's being said.

Scan for sentiment emojis

Sentiment emojis help you quickly scan a transcript so you can listen to that part of the conversation.

For example, where you see red emojis for customer turns and then a green emoji, you might choose the timestamp to jump to that specific point of the recording to hear how that agent helped the customer.

Scan for highlighted issues

Issues (also known as call drivers) are turns where the primary call driver or reason for the customer outreach is underlined in the transcript. You can use this information to identify common emerging patterns across customer conversations.

Tap or click category tags to navigate through transcript

When you tap or click on the category tags, Contact Lens auto-navigates to the corresponding point-of-interests in the transcript. There are also category markers in the recording playback visualization to indicate which part of the audio file has utterances related to the category.