Dropzone for Zendesk Messaging + AI agents: Ultimate AI (Zendesk AI Agents - Advanced) – Online Help Center

Overview

SendSafely's Dropzone for Zendesk Messenger can be used by Zendesk's Ultimate AI agent (also known as AI Agents - Advanced). The Dropzone for Zendesk Messenger is a Zendesk Conversation Extension that allows agents to request sensitive customer data and secure file uploads.

This article will walk you through configuring the Dropzone for Zendesk Messenger in Ultimate AI and provides an example that inserts the Dropzone's secure upload button when the AI agent determines a sensitive file must be collected.

Prerequisites

The article assumes Ultimate AI is already connected to Zendesk Messenger via Sunshine Conversations, as outlined in this article.
This integration requires the installation of our Dropzone Connector for Zendesk Messenger, deployed as a self-hosted AWS Lambda function. You must have access to an AWS environment with Admin rights to deploy the Lambda via cloud formation template.

Step 1: Deploy the Dropzone Connector

Follow the instructions documented in this article to deploy the Dropzone Connector for Zendesk Messenger in AWS.

Step 2: Configure API Integrations in Ultimate AI

This example requires you to configure two API Integrations in Ultimate AI:

Insert SendSafely Dropzone - Inserts the SendSafely Dropzone into a Messenger conversation
Get SendSafely Link - inserts a SendSafely package link into a conversation after the user uploads files.

Instructions for configuring both are outlined below.

2a. Insert SendSafely Dropzone

In Ultimate AI, navigate to API Integrations in the left hand menu. Next, click the blue "Add integration" in the top right corner.

1. On the Details tab provide the following:

API integration name: Insert SendSafely Dropzone
Description: Insert the SendSafely Dropzone into the conversation.

2. On the Request parameters tab, provide the following:

Key: platformConversationId
- Note: parameter platformConversationId must be named exactly as written and is case sensitive
Type: String
Test value: A valid Platform Conversation Id from a previous conversation (to obtain a sample value, go to the Ultimate AI Conversation Logs and expand the details for any existing conversation).

Click + Add parameter and add the following:

Key: email
- Note: parameter email must be named exactly as written and is case sensitive
Type: String
Test value: user@example.com

3. On the Production (Default) tab provide the following:

Environment name: Production
Method type: POST
URL: The Webhook URL copied from the AWS Stack "Outputs" tab when installing the Dropzone Connector (refer to step 6 of this article).

On the Body tab paste the following JSON.

NOTE: The value specified for the chatBotAppId is the "App ID" associated with the Zendesk Messenger API key used by your Zendesk Dropzone Connector (refer to step 7 of this article).

{
 "chatBotConversationEmail": "{{email}}",
 "chatBotConversationId": "{{platformConversationId}}",
 "chatBotAppId": "INSERT_YOUR_ZENDESK_MESSENGER_API_APP_ID_HERE",
 "extensionMode": "bot",
 "chatBotAction": "insert-extension",
}

4. On the Scenarios > Success tab provide the following:

Key: uploadId
- Note: parameter uploadId must be named exactly as written and is case sensitive
Query: data.uploadId
- Note: query data.uploadId must be named exactly as written and is case sensitive

The 'Insert SendSafely Dropzone' API integration is now configured. Test the integration to confirm it works. Once the test is successful, configure the 'Get SendSafely Link' API integration as described below:

2b. Get SendSafely Link

On the API integrations page, click the blue "Add integration" button in the top right corner.

1. On the Details tab provide the following:

API integration name: Get SendSafely Link
Description: Inserts the SendSafely package link into the conversation after the user uploads files

2. On the Request parameters tab, provide the following:

Key: uploadId
- Note: parameter uploadId must be named exactly as written and is case sensitive
Type: String
Test value: 12345ABCDE

Click + Add parameter and add the following:

Key: platformConversationId
- Note: parameter platformConversationId must be named exactly as written and is case sensitive
Type: String
Test value: A valid Platform Conversation Id from a previous conversation (to obtain a sample value, go to the Ultimate AI Conversation Logs and expand the details for any existing conversation).

3. On the Production (Default) tab provide the following:

Environment name: Production
Method type: POST
URL: The Webhook URL copied from the AWS Stack "Outputs" tab when installing the Dropzone Connector (refer to step 6 of this article).

On the Body tab paste the following JSON

NOTE: The value specified for the chatBotAppId is the "App ID" associated with the Zendesk Messenger API key used by your Zendesk Dropzone Connector (refer to step 7 of this article).

{
 "chatBotAction": "insert-link",
 "chatBotAppId": "INSERT_YOUR_ZENDESK_MESSENGER_API_APP_ID_HERE",
 "chatBotConversationId": "{{platformConversationId}}",
 "uploadId": "{{UploadId}}"
}

4. On the Scenarios > Success tab provide the following:

Key: packageLink
- Note: parameter packageLink must be named exactly as written and is case sensitive
Query: data.packageLink
- Note: query data.packageLink must be named exactly as written and is case sensitive

The 'Get SendSafely Link' API integration is now configured. Test the integration to confirm it works. The next step is to build out a use case dialog to utilize the API Integrations.

Step 3: Add Use Case (Dialog) in Ultimate AI

This last step of the configuration is to add a Use Case (Dialog) in Ultimate AI that calls the Insert SendSafely Dropzone and Get SendSafely Link API integrations from above.

From the left hand menu click Content. Either create a new Dialog, or edit an existing use case. Use the Name and Reason for request to provide your AI agent with contextual information about when to use this dialog.

A flow for the dialog is shown below. Note that the flow includes a step to collect the end-user's email address (if not already known) so that it can be passed to the Dropzone when files are uploaded.

At this point the integration is now configured and should be working. Make sure the dialog is set to “active” and you should be able to invoke the dialog from the Zendesk Web Widget.

Articles in this section