Recently, AWS announced: Amazon Transcribe Call Analytics, a new feature to Amazon Transcribe. With this new feature, AWS customers can extract valuable insights from customer conversations, such as customer and agent sentiment, and conversation characteristics such as no-talk time, interruptions, loudness, and call speed.
At AWS re:Invent 2017, the company first launched Amazon Transcribe, which offers an automatic speech recognition service (ASR) that allows developers to add speech-to-text capabilities to their applications. After the General availability in 2018, AWS further developed the service with, among other things: the automatic editing of the Personally Identifiable Information (PII) function last year. And now another feature has been added with Transcribe Call Analytics – a machine learning (ML) powered API that allows developers to analyze call recordings to get turn-by-turn call transcripts and actionable insights to add to their call applications.
Through the transcription call analysis to consoledevelopers can choose all available call category templates and configure some logic. Then they choose a model type, a general or custom language model and specify the language settings. The next steps are providing the input location, for example an S3 bucket, choosing the appropriate AWS Identity and Access Management (IAM) role, and starting the job. Finally, developers can also start a job created through the console by calling the StartCallAnalyticsJob operation of the API or AWS CLI.
Note that, for example, the output task can be visualized in a Word document by using an available Python 3 application at GitHub.
Based on ASR implemented in Transcribe, Transcribe Call Analytics adds natural language processing (NLP) features that are specifically trained on customer conversations and optimized to provide highly accurate call transcripts and actionable insights. With a simple API call, developers can now easily add call analytics to any application and extract customer insights from conversations without building AI pipelines and training custom ML models.
Currently, Transcribe Call Analytics is available in: 21 languages in the following AWS regions: US West (Oregon), US East (N. Virginia), Canada (Central), Europe (London), Europe (Frankfurt), Asia-Pacific (Mumbai), Asia-Pacific (Singapore), Asia-Pacific (Seoul), Asia-Pacific (Tokyo), Asia-Pacific (Sydney). Additionally, the feature follows the same pricing model as Transcribe — after a 12-month free trial, AWS customers pay $0.0005 (per second) for the first 250,000 minutes per minute analyzed. More details on Amazon Transcribe pricing are available at the: pricing page.