StableN/ASyncedN/A

AI Ability Use Case: Sentiment Analysis and Case Wrap up

Updated Oct 23, 20261 viewsView on usehalo.com
  • What is Halo AI's Sentiment Analysis and Case Wrap Up?
  • How does it Work? 
  • Building Sentiment Analysis and Case Wrap Up
  • Ticket Field Setup
  • Create Runbook and AI Abilities
  • Configure the Runbook 

What is Halo AI's Sentiment Analysis and Case Wrap Up?

Sentiment analysis and case wrap up is an AI-driven automation that reviews each ticket as it closes. In seconds, it tells you how the user’s experience went and what happened along the way.

AI looks at every action on the ticket and works out how the user’s sentiment changed between opening and closure. It turns this into a numerical satisfaction score you can track and compare across tickets, and a sentiment analysis, which is a short summary of how the user felt over the lifetime of the ticket. It also produces a clear summary of the key events and actions on the ticket. Anyone reviewing tickets gets a complete overview of a ticket at a glance rather than piecing it together from long action histories. Account managers in particular can quickly see how their customers are feeling, spot tickets that ended on a low note, and walk into client conversations already informed, while spending far less time on ticket reviews.

If you have a Halo environment provisioned for internal IT and enterprise solutions teams that was deployed after 28th August 2026, then this process will be pre-built in your system and ready to use without any further configuration. You can refer to the "How does it Work" section of this guide for information on how this functionality works out of the box.

If you have a Halo environment provisioned for internal IT and enterprise solutions teams that was deployed before this date, or you have an instance of another Halo solution (PSA, CRM, CSM), then you will need to set up the process using the "Building Sentiment Analysis and Case Wrap Up" section of this guide.

How does it Work?

The process is driven by a runbook prompting an AI Ability.

By default, this process is triggered by incident tickets being closed (but this can be customised). 

In the Figure 1 example, the ticket is about be to closed 

Fig 1. Closing ticket

Upon closure, the runbook "Halo AI - Sentiment Analysis & Case Wrap Up" will run. When this runs the following process occurs:

  • All actions on the ticket are compiled and sent to an AI model.
  • An Ability prompts the AI model to score the satisfaction of the user on a scale of 1-10 as well as provide an evaluation of the user's current emotional state (at the time of closing the ticket) based on the actions provided. 
  • Another AI Ability prompts the AI model to summarise the ticket based on the actions provided. 
  • The ticket is updated with the AI's response so this is visible to agents. 

Using the Figure 2 example, now the ticket is closed and the following fields have been updated:

  • AI Satisfaction Level
  • AI Sentiment Analysis
  • Case Summary

Fig 2. Ticket updated with sentiment analysis and case summary

Building Sentiment Analysis and Case Wrap Up

The sentiment analysis and case wrap up automation requires the following to be built in your instance:

The following sections of this guide will detail how to build this. Keep in mind that instances of Halo provisioned for internal IT and enterprise solutions teams, deployed after August 28th 2026 will have this built in already.

Ticket Field Setup

Create Custom Field

A custom field will need to be created to store the case summary AI generates for you. 

Head to Configuration > Custom Objects > Custom Fields > New > Name the field accordingly and ensure the "Type of Field" is set to "Text", then save. 

Fig 3. Field to store case summary

Once you have saved your custom field, make note of the field ID (visible in the URL) as this will be needed later.

Add Fields to Ticket Type

Now, you will need to add the following fields to the field lists of the ticket types you intend to run this automation on:

  • AI Satisfaction Level
  • AI Sentiment Analysis 
  • Your Case summary custom field 

Fig 4. Relevant Fields added to Ticket Type

Create Runbook and AI Abilities

The runbook and AI abilities required can either be:

  • Created in your instance manually
  • Downloaded from the online repository (from v2.250+)

Downloading from the repository is a much faster method of setup. 

We detail both methods of setup in this guide; you only need to follow one of the respective sections. 

Creation by Downloading from the Online Repository (v2.250+)

The runbook online repository can be found under Configuration > Integrations > Custom Integrations > Integration Runbooks > select the Globe symbol. 

Fig 5. Runbook Online Repository

From here select "Halo AI - Sentiment Analysis & Case Wrap Up". Then use the "Install to Halo" button to add this to your instance (this will be disabled when added). 

Note: Some instances of Halo may already have the runbook and ability required added. If this is the case an "Update local version" button will be available in place of the "Install to Halo" button, you do not need to complete any further setup. 

Fig 6. Download Runbook from the Online Repository

Adding this will create both the runbook and the required AI Abilities in your instance. 

Once added head to Configuration > Integrations > Custom Integrations > Integration Runbooks, to open up the runbook. 

Some additional configuration will need to be made to the runbook, so head to the "Configure the Runbook" section of this guide. 

Manual Creation

Create Sentiment Analysis Ability 

To create the sentiment analysis AI Ability, head to Configuration > AI > Feature Library > AI Abilities > New. Name the Ability accordingly and choose which AI connection to use. 

The instructions required for this ability can be found attached to this guide "AI Ability - Sentiment Analysis". If you are viewing this guide on our website, you will need to obtain this attachment from our portal here. 

Enable "Force JSON response" then add the input variable shown in Figure 7. Ensure you copy the details of the input variable exactly (including the description). 

Fig 7. Sentiment Analysis Ability

Create Case Wrap Up Ability 

To create the case wrap up AI Ability, head to Configuration > AI > Feature Library > AI Abilities > New. Name the Ability accordingly and choose which AI connection to use. 

The instructions required for this ability can be found attached to this guide "AI Ability - Case Wrap Up". If you are viewing this guide on our website, you will need to obtain this attachment from our portal here. 

Disable "Force JSON response" then add the input variable shown in Figure 8. Ensure you copy the details of the input variable exactly (including the description). 

Fig 8. Case Wrap Up Ability 

Create the Runbook

To create the runbook required, head to Configuration > Integrations > Custom Integrations > Integration Runbooks > Import from JSON.

Fig 9. Import Runbook using JSON

Here, paste the JSON attached to this guide, titled "Runbook JSON - Sentiment Analysis and Case Wrap Up". If you are viewing this guide on our website, you will need to obtain this attachment from our portal here. 

Once saved, click into the runbook that has been created (this will be disabled upon creation).

Head to the "Flow Chart" tab and edit the "Emotional Evaluation" step. 

Fig 10. Edit Emotional Evaluation Step

In this step, check the AI Ability for Sentiment Analysis you created earlier is set here, if it is not, set this here and save the step.

Fig 11. Sentiment Analysis Ability assigned to runbook step 

Now repeat this process for the "Case Summarisation" step, ensuring the case wrap up ability you created earlier is assigned here.

Configure the Runbook

Regardless of whether you created the required runbook manually or downloaded this from the online repository, before the automation can be used, some properties of the runbook will need to be set.

Set Which Field is Updated with the Case Summary

Under the "Flow Chart" tab edit the "Update Case Summarisation" step. 

Fig 12. Edit Case Summarisation step 

Here replace the value in the JSON body highlighted in Figure 13 with the ID of the custom field you created earlier. 

Fig 13. Value in JSON body to update

Set when the Runbook will Trigger

Now you will need to set when the runbook will trigger (run) by adding an 'Event' under the "Details" tab. 

Fig 14. Add Event to trigger runbook

We recommend adding the event "Closed" to have the runbook trigger every time a ticket is closed. However, additional conditions can be added here. 

Fig 15. Runbook set to trigger every time a ticket is closed 

Now, just enable the runbook and the automation is ready to use. 

 

Fig 16. Enable the Runbook 

Referenced by