Posted On: May 15, 2020

Amazon Chime Voice Connector now supports real-time audio streaming for media forked from Cisco Unified Border Element (CUBE) using Network-Based Recording. This makes it easy for you to build machine learning, analytics, and processing applications using real-time call audio from your on-premises Cisco phone systems. Previously, you had to integrate and deploy expensive on-premises hardware and software, or make disruptive changes to your enterprise voice network.  

With this feature, you can use Amazon Chime Voice Connector to ingest audio from phone calls though on-premises CUBE and send it to Amazon Kinesis Video Streams. It also publishes call events with metadata to Amazon EventBridge for easier correlation of audio streams with call participants across systems. You can use this feature with Amazon Machine Learning (Amazon ML) services to build applications that assist with compliance and quality assurance. For example, contact center customers can build AI-powered applications with Amazon Chime Voice Connector to improve customer experiences and enhance agent productivity. These applications can use Amazon Transcribe to transcribe the audio stream in real-time so agents can focus on callers rather than taking notes. You can also use Amazon Comprehend to provide curated suggestions to allow agents to respond faster and consistently help callers, even in unfamiliar situations.

To get started with the audio streaming for CUBE Network-Based Recording feature, visit Voice Connector page on Amazon Chime console and enable audio streaming. To learn more about the feature, visit our technical documentation and step-by-step example of an AI-powered speech analytics application. Streaming for CUBE Network-Based Recording feature is generally available starting today in the US East (N. Virginia) and US West (Oregon) regions. The feature is priced at $.0028 per minute of use. To learn more about Amazon Chime Voice Connector, see the Amazon Chime product page.