Freshworks Knowledge Ingestion

Overview

Fresh Connector supports the ingestion of Users and User Profiles. This page describes the steps and configurations needed to ingest these entities from Fresh’s external system. Keep in mind the most basic distinction in Fresh: Agents and Requesters.

By default, a new Fresh Data Source is configured to fetch Requesters.

• User Learning

  • Requesters

  • Agents

    • User Field Mappings for Agents

• User Profile Learning

  • Requesters

  • Agents

• Knowledge Learning

  • Ingest all the KBs from Freshworks

Create an Aisera Service Account

Use the Aisera Admin UI to create an Aisera Service Account User that can log into your FreshWorks system. This user only needs Read permissions (with Export ability) to transfer data to the Aisera platform DB. If you plan to use Ticket Concierge, Knowledge Generation, or other features that write back to your FreshWorks system, this user will need Read/Write permission (with Import/Export ability). This user does not need Execute or Delete permissions because all Aisera operations will be performed, tracked, and logged in the Aisera cloud.

Create an Integration

1. Go to Settings> Integration > New Integration.

2. Provide the following details:

   1. Name: Name of the Aisera Service Account User.

   2. Endpoint: Provide the URL you need to connect to.

   3. Public: Determines whether the integration can be used outside the firewall.

   4. Description: Enter the description (Optional).

3. Click Next.

4. Enter the Authentication details.

There is only one type of Authentication supported - Basic.

1. Basic.

Enter User Name and Password.

1. Click OK.

Create a Freshworks DataSource

1. Click Settings > Data Source.

2. Click + New Data Source and search for Freshworks.

3. Select the Freshworks Icon and click Next.

4. Enter the General Details:

Name	Description
Name	Name of the Data Source
Type	Select the Type: Downstream or Upstream
Integration	Select the Integration created in the step above
Functions	Select one or more of the “Functions” for which this Data Source will be used i.e Knowledge Base Learning, User Learning..etc
Schedule	Select the schedule of the DS you want it to run.
Description	Optional Description can be added

1. Enter the Configuration details:

   1. Event Type: N/A

   2. Bypass Test Connection: Disable if you want to bypass the test connection

   3. Custom Query: Enter the valid Freshworks query. I.e. the filter that you want to apply interest only to a specific set of records.

   4. Transformation Script: Enter the transformation script if you want to manipulate or transform the data being ingested from the Fresh. For more details refer to this link (Aisera internal only)

2. Click Next.

3. Override section is covered in the following sections to ingest Agents. (Other configuration details are not needed)

4. Go to the Data Source you just created and Click Play to Run the DS as shown below.

User Learning

Requesters

When creating a Fresh Datasource with the “Learn Users” function, by default it is configured to fetch Requesters from the external system and map them into User entities. The mappings are loaded by default and are ready to be used.

Agents

In order to fetch Agents from the external system, we need to configure the Override Configuration field in the Datasource config (Overrides section is visible after using the ?system parameter in the url) with this json string:

Click the below to view the JSON:

https://docs.google.com/document/d/1HvXamTDGkSKi3GN_FZ3DoUeLOzMKfSb-DMH5jXOaSzo/edit?usp=sharing

User Field Mappings for Agents

The email-related field mapping needs to be changed for Agents to “email” (from “primary_email”)

User Profile Learning

To ingest User Profiles, we also need to ingest Users. This means that we must choose both “Learn Users” and “Learn User Profiles”. Choosing only the latter will prevent the entries from being properly stored in the database.

Requesters

Fresh Data Sources are configured by default to ingest Requesters. However, there is a specific field mapping that needs to be deleted:

This field mapping (userProfileRoleName) is designed to indicate that a User Learning entry is an agent. You can either change the fixed value or delete it if it is not needed. These are the proposed field mappings for requester user profiles:

Agents

To ingest Agents as User Profiled, we need to use the same JSON Override Configuration as we did for Agent User entries:

Click the below link to View the Json:

https://docs.google.com/document/d/1uYM5210b9Xvf-oCHf6NX3yKmnCFg-I940LZstGqG7BI/edit?usp=sharing

We also need to change the field mappings that reference the email as an external field path. Now “primary_email” needs to be changed into “email”

These are the proposed field mappings for Agent User Profile entries:

Learn KBs with ACL Tags

Documents in Freshworks can contain various attributes that are related to the ACL mechanism.

By default, the Freshworks connector does not ingest tags connected to visibility. We need to use the following Override configs in a Datasource configuration:

Override Configuration:

Click the below link to View the Json:

https://docs.google.com/document/d/1J-Dc9qVCeR9MvChJ_3DJnMYrZlw6h03CG_5qfzjWF-c/edit?usp=sharing

Override Document Transformation:

Click the below link to View the Json:

https://docs.google.com/document/d/1Ju9qZ6VfsyLXvNpRu1USbprKqq7-xMko88npOLzRzTc/edit?usp=sharing

Review Crawled Knowledge Documents

After the DataSource run is completed, the administrator may review all the knowledge documents and ingested sections.

To view all the ingested documents:

1. Navigate to AI Workbench on the left-hand main navigation panel.

2. Click the Review > Knowledge Review tab at the top of the page, as shown below:

3. You will see Commit Reviewed and the number of documents that are ingested.

4. Since this is the first crawl, all the documents are in Reviewed status by default.

5. Click on Commit Reviewed and click Commit.

6. Now, click on the Knowledge tab in the left panel and you will see all the ingested documents. (It might take a couple of minutes to show/load the docs in the Knowledge tab)