AWS re:Invent 2020: Machine Learning is the Future of Contact Center

Moshe Beauford

So many updates, here's what we found out so far

AWS re:Invent 2020: Machine Learning is the Future of Contact Center

For 14 years, Amazon Web Services has dominated the cloud market, the pandemic has only accelerated that notion. That’s clear from the numbers, as AWS today extends over 175 computing, storage, networking, analytics, machine learning and artificial intelligence (AI), Internet of Things (IoT), mobile, security, hybrid, virtual and augmented reality (VR and AR), functionalities.

This week at AWS re:Invent, Amazon Web Services, Inc. dug its feet deeper into the sand, announcing a wide array of machine learning (ML) and artificial intelligence (AI), capabilities to its popular Amazon Connect contact center offering. “We want to insert machine learning into every process to enhance the human experience,” Eron Kelly, GM, Compute Infrastructure and Applications Marketing, AWS, said in an interview with UC Today.

Amazon Connect is a solution that seems to have been created for pandemics and enabling remote working on a massive scale. It is easy to use, quick to deploy, and scalable (from the smallest of contact centers to tens of thousands of agents). According to Kelly, all the recent additions to the platform give agents access to the right information at the right time, provide a more personalized service, and aim to help managers impact customer interactions in real-time.

Knowledge Base for Agents

Now available in preview, Wisdom lets agents access the information they need to solve issues in real-time. It aims to cut data silos, so contact center agents can find a range of products as well as answers to FAQs using machine learning to make the process more efficient for agents. The idea is that customers expect answers in an instant. Having access to a knowledge base that is both wide-ranging and intuitive, could work to increase agent productivity.

Here is where some of the real-time magic happens. The feature leverages Natural Language Processing (NLP) to recognize customer concerns during a call and then shows the relevant content stored in the knowledge base to agents. I also learned that there are several pre-built connectors to Salesforce and ServiceNow so agents can access relevant CRM information.

Amazon Connect Customer Profiles

Unifying the customer profile is essential if you want to reduce things like call time. Amazon Connect Customer Profiles hopes to paint a more accurate photograph of customers for agents. There, agents can find content about customer activity and experiences, including product and service information. No more third-party services and no more disparate databases with this feature, meaning agents no longer have to toggle between systems to locate what should be in one platform.

As soon as a customer calls, the Amazon Connect Customer scans and matches a customer’s records across many applications using criteria like phone numbers and account IDs. It searches transcripts, past customer sentiment, CRM software, e-commerce, and order management applications to deliver the experience. In a statement, a company spokesperson said:

“AWS customers can connect to other homegrown applications using Amazon Connect’s Software Development Kit (SDK) and Application Programming Interfaces (APIs) as well as use pre-built connectors to third-party applications like Marketo, Salesforce, ServiceNow, and Zendesk from the Amazon Connect console”

Real-Time Analytics in Amazon Contact Lens

Demonstrating the breadth and depth of Amazon’s technology, Amazon Connect added real-time analytics with a feature called Real-Time Contact Lens for Amazon Connect. Last year at re:Invent, AWS launched Contact Lens for Amazon Connect, then the offering focused on analyzing call transcripts after interactions with customers. It would give calls a positive or negative score, detect keywords, phrases, etc., all in hopes of enhancing the customer and agent experience.

Eron Kelly

Eron Kelly

Today, Kelly said, agents also want the ability to impact customer interactions during live calls to prevent poor experiences. This is where Real-Time Contact Lens comes in, hoping to make it easier for managers to make real-time adjustments to CX based on live data analysis, powered by – you guessed it – machine learning.

Contact center managers can create rules to flag customer issues using words they identify, like not happy, poor product quality, and cancel. They can leverage sentiment analysis to identify negative sentiment, voices being raised, and more. When this happens, events get flagged so managers can give input into how to proceed.

Automating Tasks with Amazon Connect

Available now, Amazon Connect Tasks can automate, track, and manage tasks for contact center agents. By AWS estimations, they say the feature works to improve agent productivity by up to 30 percent.

“Follow up is one of the areas where agents spend their most time, completing tasks for customers. They often spend their time writing notes by hand and trying to remember what they needed to do”

Kelly said that this could be sending out signature reminders, insurance claims follow-ups, etc. With Amazon Connect Tasks, companies can access pre-built connectors to CRM applications as well as APIs to integrate with their applications. Contact center managers can assign and rank tasks to agents and even do so based on their availability or skill set. No more wallboards? Not a problem, Kelly said, adding that tasks are displayed for agents in the same Amazon Connect interface they use for all interactions. Agents can create their own tasks and even automate tasks like telling a customer when their flight gets booked.

Now available in preview, is Amazon Connect Voice ID, which uses machine learning to analyze and identify customer voices. Asking their date of birth, their mother’s favorite color, and other questions only they would know the answer to, a thing of the past. Agents just know who they are when they connect for the call.

Throughout the interview, Kelly stressed to me the importance of ML and AI in the contact center and said that it would be a push for the cloud giant moving into 2021. AWS has had a solid year thanks to the pandemic and next year should be even more fruitful for the company which expects to add even more users to its growing customer base. As for Chime, we can expect the platform to keep gaining traction among those who want to enhance their apps with video and audio calling.


Join our Weekly Newsletter