No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Better for You, Better for Customers: The AI/ML-Powered Contact Center


AI/machine learning concept drawing
Image: ryzhi -
The contact center is a critical hub in any business that works to satisfy customers, not only because it’s the key to delivering satisfaction for them, but because it’s a direct window into the way they feel about the company and what it offers. However, some aspects of a contact center’s traditional operation can frustrate customers and agents alike. Think of repetitive “mother’s maiden name” queries to authenticate identity… waits, holds, or transfers to retrieve needed information… or just an emotional state that finds two people talking past each other instead of to each other.
At AWS we work to help organizations address these challenges by embedding artificial intelligence (AI) and machine learning (ML) into contact center operations. If AI and ML are still just buzzwords in your business, it’s time to put them to practical use. Here are some ways Amazon Connect, our omnichannel cloud contact center solution, can help you start:
  • Customer authentication without the friction — How many times have you had to repeat your name and identifying information over and over again during the same call? AI/ML-powered voice identification can dramatically streamline this process, using smart assistant voice technology similar to what many people use in their homes. It establishes each customer’s unique digital voiceprint, then recognizes it at the beginning of each subsequent call. Because the customer is authenticated instantly and seamlessly, the real conversation with the agent can commence without delay. It’s even possible to recognize the customer’s issue and route the call automatically, or to support IVR calls that don’t involve a live agent at all.
  • Answers on demand — When people have questions, a traditional system may put many data sources at an agent’s fingertips. But that doesn’t feel convenient when a customer has to wait for an agent to search them. Using ML-based speech analytics, Amazon Connect uses the customer’s own language to search instantly across connected repositories of relevant information, including both internal and third-party sources. The answers surface in front of the agent as they’re needed.
  • Reading the room — Is the caller content, confused, or upset? Is there a recurring issue with a specific product? Asking an agent to assess each customer’s mood and report those consolidated impressions up the chain of command would make their job even harder than it already is. Using Amazon Connect’s natural language processing (NLP) capabilities, a contact center can transcribe calls, track customer sentiment through tone of voice, detect common issues and customer trends, pinpoint discrepancies, or flag phrases like “not happy” or “cancel my subscription.” This way, the company gathers a rich array of data while the contact center agent stays focused on the matter at hand.
These are some examples of the AI/ML capabilities we’ve built directly into Amazon Connect. In dozens of languages, in more than 30 countries, the contact centers we support are seeing better responsiveness alongside reductions in call volume, call time, training time, and supervisory burdens. As we continue to innovate with AI/ML, we know these experiences are paving the way for equally valuable applications in other areas of work.
Learn more about Amazon Connect here.