Integrating Contact Flows with Agent Desktop Admin Lambda Functions

When VoiceFoundry deploys an Agent Desktop environment, several lambda functions are deployed as part of the solution to allow Contact Flows to interface with the Agent Desktop administration functions.

The Lambda Function names listed below are accurate however the environment value will change based on the values used during the deployment by the VoiceFoundry Cloud team.

As the Lambda functions reside in the VoiceFoundry Cloud account, when using the Lambda Function Invoke block in a contact flow, you must use the full ARN as per the below image.

Speak with your VoiceFoundry Customer Success Manager or email the Product team on cloudops@voicefoundry.cloud

Read the details relating to the Prompts Lambda Function in conjunction with this section to understand prompt locales.

In the below example, we have created a Banner called “Emergency”.

The below parameters are required to be used with the Banner & Events Lambda function

CODE

    "Parameters": {
    "lookupValue": "All",
    "locale": "en-AU"
    }

The Function input parameters that can be used are:

Value	Example	Use
All	All	We use the below example in our contact Flow
QueueName	BasicQueue	We use the below example in our contact Flow
DialledNumber	+11234567890	We use the below example in our contact Flow

If a matching banner is found, then the below response is received by the lambda function.

JSON

{
  "lambdaResult": "success",
  "script": "<speak>Due to an Emergency, we are temporarily closed.</speak>",
  "emergencyType": "EMERGENCY",
  "destination": "CONTINUE"
}

If no matching banner is found, then the below response is received.

JSON

{
  "lambdaResult": "fail",
  "script": "",
  "emergencyType": "",
  "destination": ""
}

If we were to pass the below request to the lambda, note the change of the locale to en-ZA

CODE

    "Parameters": {
    "lookupValue": "All",
    "locale": "en-ZA"
    }

then we would get a response

CODE

{
  "lambdaResult": "success",
  "script": "<speak>Due to an Emergency, we are temporarily closed in Africans</speak>",
  "emergencyType": "EMERGENCY",
  "destination": "CONTINUE"
}

Prompts Lambda

In the below example, we have created a prompt called “Emergency Prompt”. This prompt has two locales, which give the user the ability to change the response. The local value shows in the bracket for the language selected, which shows as en-AU

In my contact flow, when I want to return the prompt, we would invoke the lambda function and pass the function input values..

CODE

    "Parameters": {
      "promptName": "Emergency Prompt",
      "locale": "en-ZA"
    }

Example (refer to rows 32 & 33)

JSON

{
  "Name": "ContactFlowEvent",
  "Details": {
    "ContactData": {
      "Attributes": {},
      "Channel": "VOICE",
      "ContactId": "5ca32fbd-8f92-46af-92a5-6b0f970f0efe",
      "CustomerEndpoint": {
        "Address": "+11234567890",
        "Type": "TELEPHONE_NUMBER"
      },
      "InitialContactId": "5ca32fbd-8f92-46af-92a5-6b0f970f0efe",
      "InitiationMethod": "API",
      "InstanceARN": "arn:aws:connect:us-east-1:123456789012:instance/9308c2a1-9bc6-4cea-8290-6c0b4a6d38fa",
      "MediaStreams": {
        "Customer": {
          "Audio": {
            "StartFragmentNumber": "91343852333181432392682062622220590765191907586",
            "StartTimestamp": "1565781909613",
            "StreamARN": "arn:aws:kinesisvideo:us-east-1:123456789012:stream/connect-contact-a3d73b84-ce0e-479a-a9dc-5637c9d30ac9/1565272947806"
          }
        }
      },
      "PreviousContactId": "5ca32fbd-8f92-46af-92a5-6b0f970f0efe",
      "Queue": null,
      "SystemEndpoint": {
        "Address": "+11234567890",
        "Type": "TELEPHONE_NUMBER"
      }
    },
    "Parameters": {
      "promptName": "Emergency Prompt",
      "locale": "en-AU"
    }
  }
}

The response back from the Lambda Function for the above example would be

JSON

{
  "lambdaResult": "success",
  "script": "<speak>Due to an Emergency, we are temporarily closed.</speak>"
}

When using Chat, the Channel value returns as CHAT which then returns the script value without speak tags.

Example (refer to row 6)

JSON

{
  "Name": "ContactFlowEvent",
  "Details": {
    "ContactData": {
      "Attributes": {},
      "Channel": "CHAT",
      "ContactId": "5ca32fbd-8f92-46af-92a5-6b0f970f0efe",
      "CustomerEndpoint": {
        "Address": "+11234567890",
        "Type": "TELEPHONE_NUMBER"
      },
      "InitialContactId": "5ca32fbd-8f92-46af-92a5-6b0f970f0efe",
      "InitiationMethod": "API",
      "InstanceARN": "arn:aws:connect:us-east-1:123456789012:instance/9308c2a1-9bc6-4cea-8290-6c0b4a6d38fa",
      "MediaStreams": {
        "Customer": {
          "Audio": {
            "StartFragmentNumber": "91343852333181432392682062622220590765191907586",
            "StartTimestamp": "1565781909613",
            "StreamARN": "arn:aws:kinesisvideo:us-east-1:123456789012:stream/connect-contact-a3d73b84-ce0e-479a-a9dc-5637c9d30ac9/1565272947806"
          }
        }
      },
      "PreviousContactId": "5ca32fbd-8f92-46af-92a5-6b0f970f0efe",
      "Queue": null,
      "SystemEndpoint": {
        "Address": "+11234567890",
        "Type": "TELEPHONE_NUMBER"
      }
    },
    "Parameters": {
      "promptName": "Emergency Prompt",
      "locale": "en-AU"
    }
  }
}

The response back from the Lambda Function for the above example is listed below. Note there are no longer any SSML tags around the string value.

CODE

{
  "lambdaResult": "success",
  "script": "Due to an Emergency, we are temporarily closed"
}

Contact Attributes Lambda

Attributes

First set up the attributes from the Attributes tab:

You can set the attribute name and type in this section.

The attribute name is unique and case sensitive

The type can be selected from one of 3 options (String, Boolean, Number).

Deleting previously set attributes will be validated against their usage in existing attribute sets. Attributes used in existing sets will not be editable and can not be deleted.

In the example test we will use later in this guide, we will use the values outlined in red below.

Attribute Sets

Once you have created your attribute, you can create or edit Attribute Sets by adding previously defined Attributes.

Select the Attributes you want to be included in the set and fill in the values.

Keep in mind that relationship between Set and Attribute is unique per attribute - once you have added an attribute to a set, you can only add locale variations in addition to the default value

For example, if you added attribute Address to a set, you will not be able to add it again, however, you can change the default value and the locale variations.

In the below example, we have created an Attribute Set called “Emergency”.

Additional locale variations can be set from this view as well by clicking on Add locale button.

Locale variations are shown below the default value if the attribute within a set

Once the Attribute Set is set up, you can invoke it in contact flow. Below is an example. Remember to refer to the build doc for the Lambda ARN

Add mandatory parameter attributeName: your selected attribute set name

Optionally you may set the locale value and get attribute values set for that locale in the selected attribute set.

Please note that if locale value for an attribute is not set it will fallback to default value of that attribute within the set.

In my contact flow, when I want to return the Attribute Set, we would invoke the lambda function and pass the function input values.

JSON

"Parameters": {
    "attributeSet":"Emergency",
    "locale":"en-AU"
}

The lambda function would then return

JSON

{
  "lambdaResult": "success",
  "global_ivrCallback": true,
  "global_stringExample": "test value"
}