Post-call analytics with real-time transcriptions
Post-call analytics is an optional feature available with real-time Call Analytics transcriptions. In addition to the standard real-time analytics insights, post-call analytics provides you with the following:
-
Action items: Lists any action items identified in the call
-
Interruptions: Measures if and when one participant cuts off the other participant midsentence
-
Issues: Provides the issues identified in the call
-
Loudness: Measures the volume at which each participant is speaking
-
Non-talk time: Measures periods of time that do not contain speech
-
Outcomes: Provides the outcome, or resolution, identified in the call
-
Talk speed: Measures the speed at which both participants are speaking
-
Talk time: Measures the amount of time (in milliseconds) each participant spoke during the call
When enabled, post-call analytics from an audio stream produces a transcript similar to a
post-call analytics from an audio file and stores it
in the Amazon S3 bucket specified in OutputLocation. Additionally,
post-call analytics records your audio stream and saves it as an audio file (WAV
format) in the same Amazon S3 bucket. If you enable redaction, a redacted transcript and
a redacted audio file are also stored in the specified Amazon S3 bucket. Enabling
post-call analytics with your audio stream produces between two and four files, as described
here:
-
If redaction is not enabled, your output files are:
-
An unredacted transcript
-
An unredacted audio file
-
-
If redaction is enabled without the unredacted option (
redacted), your output files are:-
A redacted transcript
-
A redacted audio file
-
-
If redaction is enabled with the unredacted option (
redacted_and_unredacted), your output files are:-
A redacted transcript
-
A redacted audio file
-
An unredacted transcript
-
An unredacted audio file
-
Note that if you enable post-call analytics (PostCallAnalyticsSettings)
with your request, and you're using FLAC or OPUS-OGG media, you
do not get loudnessScore in your transcript and no audio
recordings of your stream are created. Transcribe may also not be able to provide post-call analytics for
long-running audio streams lasting longer than 90 minutes.
For more information on the insights available with post-call analytics for audio streams, refer to the post-call analytics insights section.
Tip
If you enable post-call analytics with your real-time Call Analytics request, all of your
POST_CALL and REAL-TIME categories are applied to
your post-call analytics transcription.
Enabling post-call analytics
To enable post-call analytics, you must include the PostCallAnalyticsSettings
parameter in your real-time Call Analytics request. The following parameters must be included when
PostCallAnalyticsSettings is enabled:
-
OutputLocation: The Amazon S3 bucket where you want your post-call transcript stored. -
DataAccessRoleArn: The Amazon Resource Name (ARN) of the Amazon S3 role that has permissions to access the specified Amazon S3 bucket. Note that you must also use the Trust policy for real-time analytics.
If you want a redacted version of your transcript, you can include
ContentRedactionOutput or ContentRedactionType in your
request. For more information on these parameters, see StartCallAnalyticsStreamTranscription in the API Reference.
To start a real-time Call Analytics transcription with post-call analytics enabled, you can use the AWS Management Console (demo only), HTTP/2, or WebSockets. For examples, see Starting a real-time Call Analytics transcription.
Important
Currently, the AWS Management Console only offers a demo for real-time Call Analytics with pre-loaded audio examples. If you want to use your own audio, you must use the API (HTTP/2, WebSockets, or an SDK).
Example post-call analytics output
Post-call transcripts are displayed in a turn-by-turn format by segment. They include call characteristics, sentiment, call summarization, issue detection, and (optionally) PII redaction. If any of your post-call categories are a match to the audio content, these are also present in your output.
To increase accuracy and further customize your transcripts to your use case, such as including industry-specific terms, add custom vocabularies or custom language models to your Call Analytics request. To mask, remove, or tag words you don't want in your transcription results, such as profanity, add vocabulary filtering.
Here is a compiled post-call analytics output example:
{ "JobStatus": "COMPLETED", "LanguageCode": "en-US", "AccountId": "1234567890", "Channel": "VOICE", "Participants": [{ "ParticipantRole": "AGENT" }, { "ParticipantRole": "CUSTOMER" }], "SessionId": "12a3b45c-de6f-78g9-0123-45h6ab78c901", "ContentMetadata": { "Output": "Raw" } "Transcript": [{ "LoudnessScores": [ 78.63, 78.37, 77.98, 74.18 ], "Content": "[PII], my name is [PII], how can I help?", ... "Content": "Well, I would like to cancel my recipe subscription.", "IssuesDetected": [{ "CharacterOffsets": { "Begin": 7, "End": 51 } }], ... "Content": "That's very sad to hear. Can I offer you a 50% discount to have you stay with us?", "Id": "649afe93-1e59-4ae9-a3ba-a0a613868f5d", "BeginOffsetMillis": 12180, "EndOffsetMillis": 16960, "Sentiment": "NEGATIVE", "ParticipantRole": "AGENT" }, { "LoudnessScores": [ 80.22, 79.48, 82.81 ], "Content": "That is a very generous offer. And I accept.", "Id": "f9266cba-34df-4ca8-9cea-4f62a52a7981", "BeginOffsetMillis": 17140, "EndOffsetMillis": 19860, "Sentiment": "POSITIVE", "ParticipantRole": "CUSTOMER" }, ... "Content": "Wonderful. I made all changes to your account and now this discount is applied, please check.", "OutcomesDetected": [{ "CharacterOffsets": { "Begin": 12, "End": 78 } }], ... "Content": "I will send an email with all the details to you today, and I will call you back next week to follow up. Have a wonderful evening.", "Id": "78cd0923-cafd-44a5-a66e-09515796572f", "BeginOffsetMillis": 31800, "EndOffsetMillis": 39450, "Sentiment": "POSITIVE", "ParticipantRole": "AGENT" }, { "LoudnessScores": [ 78.54, 68.76, 67.76 ], "Content": "Thank you very much, sir. Goodbye.", "Id": "5c5e6be0-8349-4767-8447-986f995af7c3", "BeginOffsetMillis": 40040, "EndOffsetMillis": 42460, "Sentiment": "POSITIVE", "ParticipantRole": "CUSTOMER" } ], ... "Categories": { "MatchedDetails": { "positive-resolution": { "PointsOfInterest": [{ "BeginOffsetMillis": 40040, "EndOffsetMillis": 42460 }] } }, "MatchedCategories": [ "positive-resolution" ] }, ... "ConversationCharacteristics": { "NonTalkTime": { "Instances": [], "TotalTimeMillis": 0 }, "Interruptions": { "TotalCount": 2, "TotalTimeMillis": 10700, "InterruptionsByInterrupter": { "AGENT": [{ "BeginOffsetMillis": 26040, "DurationMillis": 5510, "EndOffsetMillis": 31550 }], "CUSTOMER": [{ "BeginOffsetMillis": 770, "DurationMillis": 5190, "EndOffsetMillis": 5960 }] } }, "TotalConversationDurationMillis": 42460, "Sentiment": { "OverallSentiment": { "AGENT": 2.5, "CUSTOMER": 2.1 }, "SentimentByPeriod": { "QUARTER": { "AGENT": [{ "Score": 0.0, "BeginOffsetMillis": 0, "EndOffsetMillis": 9862 }, { "Score": -5.0, "BeginOffsetMillis": 9862, "EndOffsetMillis": 19725 }, { "Score": 5.0, "BeginOffsetMillis": 19725, "EndOffsetMillis": 29587 }, { "Score": 5.0, "BeginOffsetMillis": 29587, "EndOffsetMillis": 39450 } ], "CUSTOMER": [{ "Score": -2.5, "BeginOffsetMillis": 0, "EndOffsetMillis": 10615 }, { "Score": 5.0, "BeginOffsetMillis": 10615, "EndOffsetMillis": 21230 }, { "Score": 2.5, "BeginOffsetMillis": 21230, "EndOffsetMillis": 31845 }, { "Score": 5.0, "BeginOffsetMillis": 31845, "EndOffsetMillis": 42460 } ] } } }, "TalkSpeed": { "DetailsByParticipant": { "AGENT": { "AverageWordsPerMinute": 150 }, "CUSTOMER": { "AverageWordsPerMinute": 167 } } }, "TalkTime": { "DetailsByParticipant": { "AGENT": { "TotalTimeMillis": 32750 }, "CUSTOMER": { "TotalTimeMillis": 18010 } }, "TotalTimeMillis": 50760 } },...}
