.Reduce manual Call-Center efforts with Amazon Connect

Jul 26th 2022-4 min read

For everyone who isn’t aware of it: Yes, there is an AWS Service which enables you to setup a whole Call center in the cloud, Amazon Connect.

What is Amazon Connect?

Amazon Connect manages a network of telephony providers from around the world, removing the need for you to manage multiple vendors, negotiate complex multi-year contracts, or commit to peak call volumes. Its telephony service includes direct inward dial (DID) and toll-free phone numbers in 20+ countries worldwide.

What will be build?

Blog Content

Given the scenario that a customer calls the Customer Support hotline to cancel a booking. In the “old days” the customer has first to wait a decent amount of time in the waiting queue while listening to 8-bit music before being connected to a real, human speaking Call center agent. As soon as the customer and agent are connected together the agent will ask the customer for the booking number and a short verification question to make sure that the booking belongs to the customer on the other site of the phone. Now as the booking is verified the agent can go to the internal sub-system and presses a button to cancel the booking on behalf of the customer. That’s it. Easy, isn’t it?

Easy? Definitely! Efficient? Not at all. No one likes (inefficient) repetitive tasks, especially if it’s just a click on a button in a sub-system on behalf of the customer. So why not automate it?

With Amazon Connect we can build a so called “Contact flow” which describes the above mentioned workflow of the “old days”. The good thing is that an agent isn’t involved as long as there is no need for it, this enables the agent to focus on more important things of the daily doing.

Blog Content

In the figure above you can see the Contact flow of Amazon Connect which is built with a live editor directly in the Amazon Connect console. Here is a short explanation of the steps to get a rough understanding what happens behind the scenes:

  • (1): The customer will hear a warm welcome message (usage of “Play prompt” contact block)
  • (2): The customer will be asked for the intent of the call, Press 1 for a cancellation request, Press 2 for another reason (usage of ”Get customer input” contact block)
  • (3) The customer will be asked to provide the booking number via the dial pad (usage of ”Store customer input” and “Set contact attributes” contact blocks)
  • (4) The customer will be asked to provide the zip code of the payer’s address for validation purpose (usage of ”Store customer input” and “Set contact attributes” contact blocks)
  • (5) Invoking an AWS Lambda to start the cancellation process (usage of “Invoke AWS Lambda function” contact block)
  • (6) The customer will hear a confirmation message that the cancellation request will be handled (usage of “Play prompt” contact block)

⚠️ Remark ⚠️

We chose a simple Contact flow to visualize it as easy as possible but this Contact flow can be highly customized and even more optimized, the validation could for example be done step by step as the customer is still in the line to enable the customer to re-enter the informations in case of an unverified booking and to further decrease the manual efforts of the agents. In addition to that we could change the verification method to use Amazon Connect Voice ID to provide real-time caller authentication and fraud risk detection to make the interaction faster and more secure.

Last but not least…

In step (5) we are invoking an AWS Lambda, this Lambda is supposed to be the starting point of an AWS Step Functions workflow which could look as the following

Blog Content


Amazon Connect is a great AWS Service to automate a lot of processes within a Call center for increasing Customer experience while decreasing manual efforts on the agent site. The full customization possibility will allow you to build everything you could imagine. We highly recommend to everyone who is re-thinking its Call center to have a look at Amazon Connect.

☎️ + 🤖 = 🚀

globaldatanetCloud Development, Optimization & Automation



  • follow globaldatanet on instagram
  • follow globaldatanet on facebook
  • follow globaldatanet on twitter
  • follow globaldatanet on linkendin
  • follow globaldatanet on twitch
  •  listen to our serverless world podcast
  • follow globaldatanet's tech rss feed
  • follow globaldatanet at github
© 2024 by globaldatanet. All Right Reserved
Your privacy is important to us!

We use cookies on our website. Some of them are essential,while others help us to improve our online offer.
You can find more information in our Privacy policy