Amazon Q in Connect for generative AI-powered self-service

Amazon Q in Connect , a feature of Amazon Connect , is a powerful tool that helps customer service agents quickly find the information they need. Instead of wasting time searching through scattered knowledge bases, wikis, and FAQs across different systems, agents can use Amazon Q in Connect to search all connected sources at once. The system works with various platforms like Salesforce and ServiceNow, as well as company wikis, FAQs, and file systems, making it easy for agents to find answers and help customers faster.

Amazon Q in Connect uses uses advanced speech analysis and natural language processing through Contact Lens to understand customer issues during live interactions. This allows it to automatically suggest relevant solutions to agents. The results have been impressive - companies using Amazon Q in Connect report approximately 15% improvements in both customer satisfaction and first-contact resolution rates, while also reducing call handling times by about 15%.

In this blog, we will guide AnyCompany through the essential steps to enable and test Amazon Q in Connect Self-Service. This configuration will help AnyCompany leverage their knowledge management, making information more accessible for their customer self-service.

For this walk through, it is assumed you have the following prerequisites:

Step 1: Use the Amazon Connect console to create an Amazon Q Domain (Assistant) and an integration with pre-configured connectors such as Salesforce, Zendesk, Microsoft SharePoint, ServiceNow, Amazon S3, etc.

Step 2: Use the Amazon Connect console to enable bot building in Amazon Connect.

In this section, we will be configuring Amazon Connect to allow the Amazon Connect admin console to access Amazon Lex using the pre-built integration so that users can easily create, edit and continuously improve conversational AI bots powered by Amazon Lex.

Step 1: Navigate to your Amazon Connect instance by selecting the instance alias from your list of Amazon Connect instances in the AWS Console.

Step 2: In the navigation pane, choose Flows, and then go to Amazon Lex Bots. Tick both check boxes Enable Lex Bot Management in Amazon Connect and Enable Bot Analytics and Transcripts in Amazon Connect

Step 3: Click on Save to save the changes.

Step 1: Navigate to Amazon Connect in the AWS Console. Select your instance alias.

Step 2: On the Overview page, select the Access URL.

Step 3: Ensure the security profile used by your admin has the following permissions under Channels and Flows -> Bots -> View, Edit, Create.

Step 4: On the navigation menu, select Routing, and then Flows. Select the Bots tab.

Step 5: On the Bots tab, click on Create bot. Enter a bot name eg. anycompany-bot and click on Create.

Step 6: On anycompany-bot, on the Configuration tab, click on Add Language and select English (US).

Step 7: On anycompany-bot menu, on the Configuration tab, toggle Amazon Q in Connect Intent to enable it. Select the Assistant ARN. Click on Confirm.

Step 8: Select Build language and wait for the bot to complete building, indicated by a green success banner.

Step 9: On anycompany-bot menu, on the Versions tab, click on Create versions, enter a meaningful description eg. Amazon Q in Connect and click on Create.

Step 10: On anycompany-bot menu, on the Alias tab, click on Create aliases, enter an alias eg., demo, enter a meaningful description eg. demo alias is used in the Amazon Q in Connect , select 1 under Choose an existing version and tick the checkbox Enable for use in flow and flow modules. Click on Create.

Step 1: On the navigation menu, select Routing, and then Flows.

Step 2: Create a new inbound contact flow to integrate Amazon Q in Connect. In the contact flow designer enter a Flow name eg., qic-self-service-flow and add the following Flow blocks:

Step 3: Configure Flow block Amazon Q Connect. Enter a Flow block name eg. Q in Connect Assistant, tick the checkbox Set manually and select your Assistant.

Step 4: Configure Flow block Get customer input. Enter a Flow block name eg. Bot, Select Amazon Lex, under Lex bot select the Lex bot anycompany-bot and alias demo, under Customer prompt Text-to-speech or chat text under eg. What can I help you with today?

Step 5: Configure Flow block Check contact attributes. Enter a Flow block name eg. Custom actions, Under Attribute to check select Lex as Namespace, Session attributes as Key and enter Tool as Session Attribute key. Under Attribute to check select Equals and enter COMPLETE under value.

Step 6: Configure Flow block Play prompt. Enter a Flow block name eg. Closing message and enter a message eg. Thank you for choosing the Amazon Q in Connect self-service. Goodbye under Text-to-speech or chat text.

Step 7: Configure Flow block Play prompt. Enter a Flow block name eg. Bot error message and enter a message eg. An error has occurred with Q in Connect Lex bot under Text-to-speech or chat text

Step 8: Connect all Flow blocks to each as it shown. Click on Publish to save and publish the Flow

You have successfully set up Amazon Q in Connect Self-Service! In the next section, we will walk through how to test your first experience using Chat.

You have Amazon Q in Connect set up, now time to try it out!

Step 1: Log in to your Amazon Connect console with your admin user that is associated with admin security profile.

Step 2: In the navigation pane, choose Home and click on Test chat.

Step 3: Under Test chat, click on Test settings and select the contact flow qic-selfservice-flow. Click on Apply

Step 4: The bot welcomes you with What can I help you with today? Enter a message eg. Hi there, my name is Abhishek. I am interested in travel credit cards. The bot replies with an answer provided by the knowledge base.

You have successfully tested your first self-service experience using Amazon Q in Connect Self-Service.

In this blog post, we guided AnyCompany through the essential steps to enable and test Amazon Q in Connect Self-Service. This configuration helped AnyCompany leverage their knowledge management, making information more accessible for their customer self-service.