J.C. Novoa demonstrates a bank's address-change workflow across phone, SMS, and voice. A caller dials an 800 number, gets routed to SMS, and Lambda orchestrates validation through the USPS API and a MySQL CRM. All interactions stream into Redshift for dashboards, and the same workflow runs on Alexa or Lex.
This quick demo provides an example of an integration between Amazon Connect and additional AWS resources to provide an omnichannel experience. The use case scenario is address change via secure channel for a bank application, let's call it. The user will call the 800 number in Amazon Connect. It will then be transferred to the secure channel via SMS. So Lambda is at the center of processing all this data, whether the client is an Alexa skill or an Alexa-based chatbot or Amazon Connect call flow. The data will flow through Lambda. It will trigger a set of Step Functions that will manage the steps of validating the phone number via an integration to a CRM system. In this case, it's just a simple database in MySQL. It will then retrieve some data from the database, look up that, uh, caller's zip code, validate that via the USPS API, pass that information to the next step of the Step Function that will then send an SMS via Twilio. The caller will click on that link, it will open up a URL which is actually an API Gateway endpoint to another Lambda function that is managing in a serverless fashion the context of the data and storing that via DynamoDB and looking up data from a MySQL database. Ultimately, all this data across all these channels will be stored and managed within the database, and at that point it can be consolidated and pushed to a Redshift database, including the contact records that are being captured natively from Amazon Connect via Kinesis. At that point, you can use something as simple as QuickSight to draw some widgets and dashboards and reports. So let's— Try to make that call.
Now we are performing an account lookup based on the phone you are calling from. Based on the number you are calling from, you call from 561-376-3742, and your zip code was validated on Tuesday, July 10th, 2018, 19:56 and 33 seconds.
Notice that we keep track of the context of the click of the event. And from here, we can keep it in the queue, transfer somewhere else, or cancel the call. What we have just done is go up to here, send a message, and now receive the URL where I'm able to manage the interaction. If I were to interact from another channel, let me delete. These and show you Lex and Alexa, we can just pretend that this information came from Amazon Connect. But we're just going to paste it here and send it. And you'll receive— you'll see that I'll receive another message once I say yes, I want to validate this. And I can do the same thing. I can click on the link, right? This is the second time that we do this. I can submit it, change I'm just gonna change my address, submit it, and I can do the same thing from an Alexa skill. I'm just gonna do it here. And paste it, and I will receive from 613,763,742, and your zip code was validated on Tuesday, July 28th, 1958, GMT+0 UTC. All this information, of course, is being managed here. You see how we just changed to the latest interaction and the latest step that was completed for that particular phone number. So I'm able to look up the information, step through all these channels by leveraging Amazon Connect, Lambda, Step Functions, Alexa, Lex, all within the AWS ecosystem. And that's it.
