Amazon Transcribe Call Analytics adds AI and ML to customer care

Get a free Techzine subscription!

Amazon Web Services is enhancing its AWS Contact Center Intelligence by adding a new transcription analytics tool designed to make insights surfaced from customer conversations significantly better.

Amazon Transcribe Call Analytics is the given name for the machine learning-powered application programming interface designed to work with the existing Amazon Transcribe tool used to generate transcriptions of customer care calls.

As Julien Simon, Amazon’s AI and ML evangelist wrote in a blog post, even the blandest calls present a chance to know your customers better, an opportunity that he believes no organization should pass on.

A better way to analyze interactions

Simon said that it is important for customer care staff to figure out why customers are calling them and measure satisfaction during the calls.

He added that any contact center offering can apply Amazon Transcribe Call Analytics since it is not limited to just AWS Contact Center Intelligence. The tool will analyze calls over the phone and web-based audio/video calls.

Before now, customers would have to make their own AI and ML models and merge them to achieve this. The tool is powered by natural language processing models trained using customer calls to make it easy to generate accurate scripts and actionable insight for each call.

Analysis made easy

Using a simple API, developers can now add analytics to any application to surface insights from conversations, eliminating the need to build AI pipelines and train customized ML models.

The insights offer information on trending issues and track performance indicators, allowing the business to gain intelligence on agent and customer sentiment, what drives the call, and details related to interruptions or hiccups in communication.

The conversations can also be tagged based on keywords to enable capabilities that include the redaction of private information in scripts.