Covering Disruptive Technology Powering Business in The Digital Age

Home > Archives > Blog > Transforming Customer Service with Amazon Transcribe Call Analytics
Transforming Customer Service with Amazon Transcribe Call Analytics
August 12, 2021 Blog


Written by: Nur Atikah Yusri, Journalist, AOPG

Analytics denotes computer data analysis to find valuable patterns easily when the data is in text form. But when businesses have audio or video data, they must find ways to convert that data for use in software and applications. The process of which requires transcribing.

Working with a transcriptionist is not a bad idea if you only needed to transcribe a single piece of audio or video. Companies who need a large number of customer service calls or meeting recordings to be transcribed, though, might find this to be costly.

The evolution of audio transcription has come a long way since humans have had to transcribe speech-to-text manually. Now, there is an abundance of applications and tools to help with transcription efforts that go beyond transcribing.

When Amazon first launched Amazon Transcribe in 2017, it acted as a cost-effective replacement for transcription providers. Using Automatic Speech Recognition (ARS), it can convert speech to text quickly and accurately. Applicable for a wide variety of uses – to transcribe customer service calls, generate subtitles, create meeting transcriptions, and transcribe medical data, among other benefits.

Now, Amazon has launched a new service that is catered to the customer service industry in the form of Amazon Transcribe Call Analytics. As an addition to Amazon Transcribe, Amazon Transcribe Call Analytics is a machine learning-powered conversation insights API that uses natural language processing (NLP) trained explicitly on customer calls to provide accurate call transcriptions and insights. The new service acts to improve the post-call analytics offerings of AWS Contact Intelligence Centre.

Here are some key features of the new Amazon Transcribe Call Analytics:

  • Detailed call analytics and conversation insights allow you to integrate actions such as sentiment, detected issues, and speech characteristics into your call analytics applications.
  • Allows you to build categories based on your needs to monitor compliance with organisational policies and procedures using automated call categorisation.
  • Provides rich call transcripts that supply your agents with insights on customers’ sentiments and issues for reference.
  • Protects sensitive customer data by redacting information such as name, address, identification, and credit card numbers from audio and text files.

These insights can help agents respond better to future customers and monitor their compliance to procedure and performance. Businesses can map trends within problems reported by customers to upgrade or look into improving services that receive many complaints. With turn-by-turn transcription, detailing more than just what is said, transcriptions can be used to devise new agent training modules complete with real-life examples. The loudness levels and scoring provided allows for a better understanding of customers’ reactions to agents’ responses.

As an API, Amazon Transcribe Call Analytics gives you the flexibility to add its capabilities to any sales or customer service programs reducing implementation time. It also does not necessitate ML proficiency of any kind as it comes with pre-trained models catered to conversational data. All API outputs are made available in an Amazon S3 bucket. Amazon provides step-by-step instructions on transcribing with call analytics using the Amazon Transcribe console, API and AWS CLI. The pricing for the service is based on a tiered pay-as-you-go plan billed monthly according to the number of seconds utilised.

Amazon Transcribe Call Analytics is transforming customer service by allowing access to analytics that can help businesses increase customer satisfaction levels. Click here to learn more.