Transforming insights to impact in your call center - Voice to action framework with Amazon Connect

The emergence of virtual telephony voice agents offers step-wise enhancements in delivering real-time insights from data, and impactful customer experiences. We explain how virtual agents can autonomously manage low-value interactions, enhance IVR-to-human agent transitions, and provide a more human-like IVR experience featuring powerful self-service capabilities. Learn how to build the solution using Amazon Connect, omnichannel cloud contact center service with real-time audio parsing via Amazon Kinesis Video Streams, enabling you to securely stream video from connected devices to AWS for analytics, machine learning playback, and other processing. The session also covers the use of algorithms for effective silence and speech detection, extraction of high-quality transcripts with Amazon SageMaker and LLM inference through Claude3 on Amazon Bedrock. We also demonstrate how to build a fast and secure LLM-powered virtual voice agent that operates in real-time and enabling secure external API access. Find out how to develop rich secure virtual agents which you can easily deploy, maintain and observe in your call centre. We also discuss the recent advances in LLM technologies and ways to reap the benefits in your voice channels.

Speakers: 
Josh Passenger, Principal Solutions Architect, AWS
Charles Prosper, Principal Applied Scientist, AWS