Present dwell agent help in your chatbot customers with Amazon Lex and Talkdesk cloud contact middle

Amazon Lex offers superior conversational synthetic intelligence (AI) capabilities to allow self-service assist in your group’s contact middle. With Amazon Lex, you may implement an omnichannel technique the place prospects have interaction through telephone, web sites, and messaging platforms. The bots can reply FAQs, present self-service experiences, or triage buyer requests earlier than transferring to a human agent. Amazon Lex integrates with state-of-the-art contact centers together with Amazon Connect, Genesys Cloud, and Amazon Chime SDK to facilitate a seamless omnichannel expertise.

That is the second publish of a two-part collection. The mixing of Amazon Lex with Talkdesk cloud contact center is impressed by WaFd Bank (WaFd)’s digital innovation journey to enhance customer experience. In our earlier post, we described how Amazon Lex integrates with the Talkdesk cloud contact middle for the voice channel. On this publish, we’re specializing in the chat channel to indicate methods to use Amazon Lex and the Amazon Lex Web UI to allow dwell brokers to work together together with your prospects in actual time. For instance, the next determine reveals screenshots of a chatbot transitioning a buyer to a dwell agent chat (courtesy of WaFd Financial institution).

Resolution overview

The next diagram illustrates the answer structure.

Within the previous structure, the next sequence of steps takes place in a dwell buyer/agent dialog:

1. Utilizing the Amazon Lex Internet UI, a buyer asks to be related to an agent. The related Amazon Lex chatbot is configured with an escalation intent to course of the incoming agent help request.
2. The Amazon Lex achievement AWS Lambda operate retrieves the Talkdesk touchpoint ID and Talkdesk OAuth secrets and techniques from AWS Secrets Manager and initiates a request to Talkdesk Digital Join utilizing the Start a Conversation API. Within the payload, the operate contains data that could be helpful to an agent, such because the buyer sentiment or the historical past of beforehand traversed intents.
3. If the request to the Talkdesk API is profitable, a Talkdesk dialog ID is returned to Amazon Lex.
4. The Amazon Lex achievement Lambda operate shops the dialog ID in Amazon Lex session attributes, thus making the dialog ID accessible to the Amazon Lex Internet UI.
5. The Amazon Lex Internet UI opens a communication session with brokers on the Talkdesk contact middle by means of a WebSocket API in Amazon API Gateway.
6. The Lambda related to the WebSocket API first shops the Talkdesk dialog ID to WebSocket consumer ID mappings in Amazon DynamoDB. Then, by means of the Talkdesk Send a Message API, the Lambda operate sends the client’s message to the agent on Talkdesk contact middle.
7. Your agent responds to the client with a message despatched by means of the callback Rest API in API Gateway. The payload contains the dialog ID of the lively dialog.
8. The callback Relaxation API is configured to assist the brokers’ incoming messages in addition to the agent’s closing of the dialog. So as to ship the agent’s message to the client, the supporting Lambda operate reads the WebSocket consumer ID related to the dialog ID from the DynamoDB desk. This makes certain the agent’s message is delivered to the suitable WebSocket consumer ID.
9. The agent’s response is displayed by means of the Amazon Lex Internet UI and the client responds or closes the chat as applicable. Steps 6–9 are repeated so long as the dialog stays lively. If the agent ends the dialog, the client is notified and the WebSocket connection is closed.

Within the following sections, we stroll you thru the steps to construct the answer structure. Dependencies amongst every step are cross-referenced.

Stipulations

To implement the answer introduced on this publish, it is best to first familiarize your self with the next AWS providers and options:

Moreover, you need to be aware of the next Talkdesk providers:

Put together your Talkdesk occasion for the Amazon Lex Internet UI chat with an agent

This part outlines the essential steps required to configure the Talkdesk chat with agent expertise utilizing the Talkdesk Digital Join channel. Overview Talkdesk APIs for additional particulars for any extra duties that could be required as a part of your particular implementation.

Full the next steps:

1. Allow Talkdesk Digital Join in your Talkdesk occasion.
2. Configure your brokers’ accounts and assign them to the brokers’ queues.
3. Construct a Talkdesk Studio flow.

This will likely be used to ship chat customers to an inbox for brokers to assign. A sample is supplied with this resolution.

4. To create an integration in your Amazon Lex Internet UI occasion, within the Talkdesk Builder navigation pane, choose Integrations.
5. On the Actions tab, configure three actions utilizing the enter and output schemas offered by means of the next hyperlinks:

6. Create a Talkdesk Digital Join Touchpoint.
7. Title the Touchpoint Lex Internet UI Chat and report the Touchpoint ID.

This will likely be saved in Secrets and techniques Supervisor as dev/talkdesk/touchpoint/ids.

8. In Talkdesk Builder, select OAuth Shoppers within the navigation pane to arrange OAuth credentials.
9. Choose Grant kind for Consumer credentials and set Scope to digital-connect:write.
10. File the consumer ID and secret key from the Keys tab.

These will likely be saved in Secrets and techniques Supervisor as dev/talkdesk/consumer/keys and used to authenticate and talk with the Talkdesk API.

11. In your AWS account, retailer the 2 secrets and techniques in Secrets and techniques Supervisor.

The next screenshot reveals the main points of the Touchpoint ID as a Secrets and techniques Supervisor secret.

The next screenshot reveals the main points of the consumer ID as a Secrets and techniques Supervisor secret.

Deploy the Talkdesk Amazon Lex CloudFormation template

The next AWS CloudFormation template creates all of the sources of the answer structure. This contains all vital IAM roles to invoke API operations, run related Lambda features, entry secrets and techniques on Secrets and techniques Supervisor, and retailer and retrieve dialog ID and WebSocket consumer ID pairs from DynamoDB.

To facilitate monitoring and debugging, a CloudWatch log group is created for every of the sources.

The CloudFormation template offers extra particulars for every of the sources.

Full the next steps to deploy the template:

1. Register to the AWS Management Console.
2. Select Launch Stack in your AWS Area to start the CloudFormation stack creation course of.
3. For Stack identify, enter a reputation.
4. For TDAUTHHOST, enter the URL of your Talkdesk occasion.
5. Depart the opposite parameters as default and select Subsequent
   
6. Choose the acknowledgement verify bins and select Create stack.
   
7. After the CloudFormation template is full, report the values for the next keys on the Outputs tab to make use of in later steps:
   • APIGatewayApiKey
   • BotAliasId
   • BotId
   • CallbackRestAPI
   • WebSocketAPIEndpoint

   

Replace the Talkdesk occasion

Log in to your Talkdesk occasion and full the next steps to replace your occasion:

1. In Talkdesk Builder, choose Integrations within the navigation pane.
2. On the Settings tab, find Base path and enter the callback Relaxation API URL you recorded earlier.
3. Underneath Different settings, set x-api-key to the worth of the API Gateway key.

Deploy the Amazon Lex Internet UI

The answer outlined on this publish makes use of the Amazon Lex Internet UI, a full-featured internet consumer to deploy your Amazon Lex chatbot in your web site. With the Amazon Lex Internet UI, you may shortly convey your chatbot-powered utility to life whereas minimizing time-to-value.

1. Select Launch Stack for the Area through which you’ll use your chatbot:
2. For LexV2BotId, enter the worth for BotId.
3. For LexV2BotAliasId, enter the worth for BotAliasId.
   
4. Launch the stack.
5. When deployment is full, find the Amazon Simple Storage Service (Amazon S3) URL for WebAppBucket.
6. Navigate to the S3 bucket on the Amazon S3 console and obtain the lex-web-ui-loader-config.json file.
   
7. Open the file and modify or add the next parameters:
   1. Within the join configuration part, add the brand new parameter talkDeskWebsocketEndpoint and set its worth to the WebSocket endpoint.
   2. Within the UI configuration part, set enableLiveChat to true.

   

8. Add the modified lex-web-ui-loader-config.json file and overwrite the earlier model of the file within the S3 bucket.
9. Return to the CloudFormation stack Outputs tab and discover the WebAppDomainName hyperlink.

This can redirect you to a full-page model of the Amazon Lex Internet UI. From right here, you may take a look at the Talkdesk integration and ensure that the bot is ready to hook up with Talkdesk utilizing the WebSocket connection.

Check the answer

Now you’re able to strive the Amazon Lex and Talkdesk chat interplay:

1. Begin your Banking Bot chat window utilizing the WebAppUrl offered as output within the CloudFormation stack.
2. Log in to your Talkdesk Digital Join channel and navigate to Conversations.
3. Within the Banking Bot chat window, request to speak to an agent.
4. Watch the client’s message being delivered to the Talkdesk Conversations Inbox.
5. The Talkdesk agent self-assigns the dialog and begins partaking with the client.

The next video demonstrates the chat expertise.

Clear up

To wash up your sources, full the next steps:

1. On the AWS CloudFormation console, choose Stacks within the navigation pane.
2. Choose the LexTalkdesk stack (or the stack identify you offered), and choose Delete.
3. Delete the stack sources by deciding on Delete stack.

Conclusion

Amazon Lex brings the facility of conversational self-service to your buyer most popular channels, reminiscent of telephone, internet chat, and messaging functions. On this publish, we demonstrated an answer that gives dwell agent help in your web site with Amazon Lex, Amazon Lex Internet UI, and Talkdesk cloud contact middle. We offered a CloudFormation stack that features DynamoDB and Lambda sources, and a Relaxation API and WebSocket API in API Gateway to take care of a communication session with brokers within the Talkdesk contact middle.

This resolution is supposed to be a reference structure or a fast implementation information that may be tailor-made to fit your group’s necessities. When you need assistance establishing this resolution, AWS Professional Services and Talkdesk can be found that will help you and your crew by means of the method of choosing the precise applied sciences in your cloud contact middle.

Concerning the authors

Grazia Russo Lassner is a Senior Guide with the AWS Skilled Providers Pure Language AI crew. She specialises in designing and growing conversational AI options utilizing AWS applied sciences for purchasers in varied industries. Exterior of labor, she enjoys seaside weekends, studying the most recent fiction books, and household time.

Austin Johnson is a Options Architect, serving to to take care of the Lex Internet UI open supply library.

Chris Brown is a Principal Pure Language AI guide at AWS targeted on digital buyer experiences – together with cell apps, web sites, advertising campaigns, and most not too long ago conversational AI functions. Chris is an award-winning strategist and product supervisor – working with the Fortune 100 to ship the perfect experiences for his or her prospects. In his free time, Chris enjoys touring, music, artwork, and experiencing new cultures.

Bruno Mateus is a Principal Engineer at Talkdesk. With over 20 years of expertise within the software program business, he specialises in large-scale distributed programs. When not working, he enjoys spending time outdoors together with his household, trekking, mountain bike using, and motorbike using.

Jonathan Diedrich is a Principal Options Guide at Talkdesk. He works on enterprise and strategic tasks to make sure technical execution and adoption. Exterior of labor, he enjoys ice hockey and video games with the household.

Crispim Tribuna is a Senior Software program Engineer at Talkdesk at the moment specializing in the AI-based digital agent mission. He has over 17 years of expertise in pc science, with a give attention to telecommunications, IPTV, and fraud prevention. In his free time, he enjoys spending time together with his household, working (he has accomplished three marathons), and using bikes.