As part of AWS re:Invent 2024, Amazon Connect introduced multiple new features including generative AI-powered customer segmentation for targeted campaigns and native WhatsApp integration for omnichannel support.
“We see WhatsApp as critical channel, but mostly outside the US,” said Michael Wallace, Solutions Architecture Leader for Customer Experience with Amazon Connect. “I think there'll still be some adoption inside the US, but globally – like in Latin America, where WhatsApp is used a lot and where SMS isn't quite as prevalent – WhatsApp will be a continued growth area for us.”
Amazon Connect also announced the integration of Gen AI-powered features into its outbound communications capabilities. With Gen AI, customers can use natural language descriptions to prompt Amazon Connect to create customer segments and trigger outbound campaigns based on specific customer events.
Wallace said that outbound campaigns are usually designed with a great deal of manual effort. The campaign manager must comb through the customer database to identify the customers they want to contact – based on demographics, region, spending level, etc. One of the less-hyped capabilities of Gen AI is its ability to parse massive dataset, find patterns and then provide results in response to a natural language prompt.
“Let's say you're a public utility and you want to contact all your customers about an upcoming, planned outage that is in a certain area, and you have some geo fencing data around that. Rather than manually sift through records yourself, you could say ‘Write me a campaign around everybody that has the postal code of 83835 and, poof, there it is,” Wallace said. “So instead of having to go through the manual process of filtering that data, [the system] will do it for you in seconds and create your multi-channel campaign.”
And based on customer data, Wallace said that their system can suggest campaigns based on actual trends, as illustrated below. The user could click on the first dialogue box and then get walked through a process that culminates in the system automatically generate a proactive campaign targeting those customers who spent more than $1,000 a month.
Amazon Q in Connect now includes customizable AI guardrails which allow organizations to block undesirable topics for self-service, filter harmful and inappropriate content based on their responsible AI policies, redact sensitive information to protect customer privacy, and verify model responses using contextual grounding checks. These safeguards can be selectively applied based on contact type, offering flexible control over AI interactions.
Amazon Connect now also includes enhancements that help contact center managers spot important trends in customer feedback and identify agent coaching insights. For example, through natural language prompts and Gen AI, managers can automatically categorize customer contacts to understand call trends over time, flag calls indicating customer discontent, learn about communication breakdowns during calls, discover agent performance improvement opportunities, and more.
Lastly, Amazon also announced other new capabilities including secure collection of sensitive customer data in chat interactions, and simplified conversational AI bot management directly within the Amazon Connect interface. Amazon Connect also added new analytics capabilities through Amazon Contact Lens to help optimize bot performance and contact center operations.
Check out these other articles on Amazon Connect:
