Keeping up with all of the new features that are constantly being released within the AWS ecosystem is next to impossible. However, keeping up with all of the Amazon announcements that affect just Amazon Connect is much more doable. That’s why we’ve rounded up the latest from Connect in the month of June 2020 and provided you with the breakdown of each.
Amazon Connect now supports higher-quality, natural-sounding Text-to-Speech voices
Amazon Connect enables customers to use Amazon Polly’s Neural Text-to-Speech Voices within their contact-center. These new voices deliver groundbreaking improvements in speech quality through a new machine learning approach, making automated conversations sound more lifelike by improving the pitch, inflection, intonation, and tempo. The new neural voices are available in eight US English, three UK English, one Spanish, and one Portuguese voice.

This feature is available in the US East (N. Virginia), US West (Oregon), and Asia Pacific (Sydney) regions and is included in the Amazon Connect service. There is no additional charge for using neural Text-to-Speech voices. To learn more, see the help documentation, or visit the Amazon Connect website.
Amazon Connect adds filtering by channel to the ‘Get queue Metrics’ block
You can now filter queue metrics by channel within contact flows for a single queue. With the launch of chat, queue metrics blended both voice and chat contacts. Now you can filter queue metrics by contact type when configuring the 'Get queue metrics' block. For example, based on the number of contacts in a blended queue, the contact flow designer can calculate the effective hold time to manage the duration a customer spends waiting.
This feature is available in all AWS regions where Amazon Connect is offered. There is no charge to use the channel filter from the Get queue metrics block from the contact flow editor, other than the Amazon Connect service usage and any associated telephony charges. To learn more, see the help documentation, or visit the Amazon Connect website.
Amazon Connect allows you to continue engaging with your customer after an agent hangs-up
You can set a contact flow after an agent disconnects from a customer call by transferring them to an automated flow. Previously, contact-center administrators did not have a native way to configure the customer experience when an agent hung up; including conducting a post-call survey, placing a customer back into a queue, or scheduling a queued call back. Now, you can define your customer interaction when an agent disconnects by adding the ‘Set disconnect flow’ block to your contact flow.

This feature is available in all AWS regions where Amazon Connect is offered. There is no charge to use ‘disconnect flows’ from within the contact flow editor other than the Amazon Connect service usage and any associated telephony charges. To learn more, see the help documentation, or visit the Amazon Connect website.
Amazon Connect now supports Kevin, Polly’s latest text to speech voice
Amazon Connect enables customers to use Amazon Polly’s latest US-English neural voice, Kevin. Built on Amazon Polly’s neural text-to-speech technology, this new voice delivers groundbreaking improvements in speech quality, making automated conversations sound more lifelike by improving the pitch, inflection, intonation, and tempo.
This feature is available in the US East (N. Virginia), US West (Oregon), and Asia Pacific (Sydney) regions and is included in the Amazon Connect service. There is no additional charge for using neural Text-to-Speech voices. To learn more, see the help documentation, or visit the Amazon Connect website.
So which one(s) should I care about?
Out of the four new features added to Connect, the most significant in our opinion is the ability to continue engaging with your customer after the agent hangs up. This is a big step in the right direction for Amazon to level the playing field with their competitors. This gives you, the contact center administrator, the ability to easily program in complex multi-step post call survey flows with very little design and implementation time. This is what we have been waiting a while for and it has come not a moment too soon.
One other honorable mention here is the introduction of Neural Voice support for Connect via Polly. These frighteningly lifelike voices can give you the ability to design your IVR to sound like a newscaster, or a book narrator for example, and small inflections such as auto-breath and prosody control (already available in Polly via SSML) have given Polly the ability to almost fool us a couple times if it weren’t for tiny nuances we’ve noticed. If you don’t believe us, go ahead and check out the newest Neural Voice release called Kevin.