Overview

Integrate Fullview AI in your Intercom Fin workflows and let Fullview AI automatically guide your customers in your product, with on-screen assistance.

To enable Fullview AI to work seamlessly with Intercom Fin, you’ll need to set up several data connectors from your Intercom workspace. These connectors allow Fin to trigger actions from Fullview AI at different points in the workflow, such as starting a session, providing feedback, or escalating to a human agent.

Prerequisites

Fullview app installed from Intercom's app store
Intercom Fin enabled for your Intercom workflows
Fullview AI enabled for your Fullview organization

1.Create Data Connectors

You can create a new data connector in Intercom via:

Settings → Data Connectors → New Data Connector

❗️Replace each connector’s request body values with the corresponding Intercom attributes. Use your Fullview organization ID for organisationId

1.1 Check if query can be handled

Used to check if the given query can be handled by Fullview

Under the About tab, set the name of the connector (eg. Check Query Confidence)

API connection tab:

Method: POST
HTTPS URL - depending on the region your Fullview org is in:

Request Body:

{

   "organisationId": "YOUR FULLVIEW ORG ID",
   "email":"Email Attribute",
   "query": "Last Message Body",
   "externalId": "User ID Attribute",
   "integrationData": {
            "platform": "intercom",
            "caseId": "Conversation ID",
            "intercomUserId": "Contact ID",
            "workspaceId": "Workspace ID"
     }

}

Response:
```
{
  "canBeAnswered": true
} 
```
- you can use this response and decide if you engage fullview or not

1.2 Start AI Session

Used to send Fullview AI the customer intent and start the assistance

Under the About tab, set the name of the connector (eg. Start Fullview AI), description and audience (eg. Customers)

API connection tab:

Method: POST
HTTPS URL - depending on the region your Fullview org is in:

Request Body:

{

   "organisationId": "YOUR FULLVIEW ORG ID",
   "email":"Email Attribute",
   "query": "Last Message Body",
   "externalId": "User ID Attribute",
   "integrationData": {
            "platform": "intercom",
             "caseId": "Conversation ID",
            "intercomUserId": "Contact ID",
            "workspaceId": "Workspace ID"
     }

}

1.3 Terminate AI Session

Used when the customer wants to end the Fullview AI assistance

Under the About tab, set the name of the connector (eg. Terminate Fullview AI), description and audience (eg. Customers)
API connection tab:
- Method: POST
- HTTPS URL - depending on the region your Fullview org is in:
- Request Body:
```
{
   "organisationId": "YOUR FULLVIEW ORG ID",
   "email":"Email Attribute",
   "externalId": "User ID Attribute",
   "integrationData": {
            "platform": "intercom"
     }
} 
```

1.4 Send Positive Feedback

Used for sending Fullview AI the customer feedback, when the session ended

Under the About tab, set the name of the connector (eg. Send Positive Feedback), description and audience (eg. Customers)
API connection tab:
- Method: POST
- HTTPS URL - depending on the region your Fullview org is in:
- Request Body - same as at the previous connector:
```
{
   "organisationId": "YOUR FULLVIEW ORG ID",
   "email":"Email",
   "externalId": "User ID",
   "caseId": "Conversation ID"
} 
```

1.5 Send Negative Feedback

Used for sending Fullview AI the negative customer feedback, when the session ended and the customer rated the interaction with a thumbs down

Under the About tab, set the name of the connector (eg. Send Negative Feedback), description and audience (eg. Customers)
API connection tab:
- Method: POST
- HTTPS URL - depending on the region your Fullview org is in:
- Request Body - same as at the previous connector:
```
{
   "organisationId": "YOUR FULLVIEW ORG ID",
   "email":"Email",
   "externalId": "User ID",
   "caseId": "Conversation ID"
} 
```

1.6 Send Negative Feedback Message

Used for sending Fullview AI the additional feedback from the customer, when the session ended and the customer gave additional feedback besides the thumbs down

Under the About tab, set the name of the connector (eg. Send Negative Feedback), description and audience (eg. Customers)
API connection tab:
- Method: POST
- HTTPS URL - depending on the region your Fullview org is in:
- Request Body - same as at the previous connector:
```
{
   "organisationId": "YOUR FULLVIEW ORG ID",
   "email":"Email",
   "externalId": "User ID",
   "caseId": "Conversation ID",
   "message": "Last Message Body"
} 
```

1.7 Note for unsuccessful resolution

This data connector will be used for unsuccessful resolutions and escalations and will create a note on the conversation in Intercom with metadata about the assistance

Under the About tab, set the name of the connector (eg. Note for unsuccessful resolution), description and audience (eg. Customers)
API connection tab:
- Method: POST
- HTTPS URL - depending on the region your Fullview org is in:
- Request Body - same as at the previous connector:
```
{

   "organisationId": "YOUR FULLVIEW ORG ID",
   "email":"Email",
   "externalId": "User ID",
   "integrationData": {
            "platform": "intercom",
            "caseId": "Conversation ID",
            "workspaceId": "Workspace ID"
     }
} 
```

Here is an example of how the configuration on a data connector should look like:

2.What to do with API data connectors?

Now that you’ve created the API data connectors, you can call them within your existing Fin workflows to pass customer support requests to Fullview AI. This enables Fullview AI to receive and process support requests submitted by customers via your Intercom Messenger, allowing for intelligent handling and automation of queries.

2.1 Examples

Here's an example of a simple workflow for a support request submitted via Intercom's chat interface and passed to Fullview AI via the Start AI Session connector.

Here's another example of how to leverage the connectors to pass Fullview AI the feedback submitted by the customer at the end of the Fullview AI assistance:

Here's another example of how to check if Fullview can handle the request.

2.2 Branches

When a customer ends a session by clicking “I’ll take it from here,” Fullview AI will automatically post one of the following messages in the conversation:

Fullview AI – Completed assistance
Fullview AI – Interrupted assistance
Fullview AI – Unsuccessful assistance

You can use these messages as triggers in your Intercom workflow by adding a Branch block that checks for the presence of any of these responses. When detected, direct the flow to request customer feedback, allowing users to rate the session (see example above).

Fullview AI <> Intercom workflow