How to Configure AI Trigger Orchestration in BoldDesk

AI Trigger Orchestration in BoldDesk enables administrators to automate ticket-handling workflows by invoking an AI Agent when a ticket meets predefined conditions. When a configured rule is triggered, the AI Agent analyzes selected ticket context fields, follows the provided instructions, and generates an output based on the ticket information.

By default, the AI Agent’s response is added to the ticket as a private note. Optionally, the response can be mapped to ticket actions, allowing BoldDesk to automatically update ticket properties such as category, priority, status, or tags.

This automation helps support teams reduce manual effort, maintain consistent ticket processing, improve response accuracy, and accelerate ticket triage. AI Trigger Orchestration can be used for a variety of scenarios, including ticket categorization, priority assignment, sentiment analysis, ticket summarization, and SLA risk detection.

Prerequisites

Before configuring AI Trigger Orchestration, ensure that the following requirements are met:

Your BoldDesk account has sufficient AI credits to run AI Agent operations.
At least one AI Agent has been created and published.
You have identified the ticket fields that should be provided to the AI Agent as contextual information.
If response mapping will be used, you have identified the ticket properties that should be updated automatically.

Each AI Trigger Orchestration execution that invokes an AI Agent may consume AI credits. Credit usage varies based on the complexity of the AI task, the volume of ticket context provided, and the response generated.

Configuring AI Trigger Orchestration

Follow these steps to create an AI Trigger Orchestration rule.

Step 1. Open AI Trigger Orchestration

Go to Admin.
In the left navigation menu, select AI Automation.
Click AI Trigger Orchestration.
Click Add to create a new orchestration rule.

From this page, you can create, edit, and manage orchestration rules.

Step 2. Choose the Rule Condition

The Choose Condition step determines when the AI Agent should be invoked.

Enter a unique Rule Name.
Optionally, provide a Description.
Configure the conditions.
Click Next.

The AI Agent is invoked only when the configured conditions are met successfully.

Step 3. Configure AI Agent Invocation

The Invoke AI Agent runs once when the rule matches and returns a response.

Choose an AI Agent

In the Choose AI Agent field, select the AI Agent that should process matching tickets from the dropdown. Optionally, click View agent to open and edit the selected AI Agent.
Verify that the selected AI Agent is appropriate for the intended workflow.

The selected AI Agent determines how the ticket information is analyzed and processed.

Select Ticket Context Fields

Ticket Context Fields determine which ticket information is sent to the AI Agent.

In the Ticket Context Fields field, choose one or more ticket fields that the AI Agent should use while generating the response.
Select only the fields that are relevant to the AI Agent’s task.

Examples of commonly used context fields include: Ticket ID, Subject, Description, Brand Name, Category, Priority, Requester Details, Custom Fields, etc. Providing relevant context helps the AI Agent generate more accurate and meaningful responses.

Configure Additional Instructions

The Additional Instructions field allows administrators to define how the AI Agent should process tickets.

Enter clear and specific instructions.
Define business rules, response expectations, or output formats.
Include any workflow-specific guidance required by your organization.

Additional Instructions work together with the instructions already configured in the selected AI Agent. They do not replace or override the AI Agent’s instructions. Use Additional Instructions to provide workflow-specific guidance, such as categorization rules, escalation criteria, or response formats, while keeping common instructions centralized in the AI Agent.
Example: A shared AI Agent can define standard response rules for all tickets, while Additional Instructions specify how tickets should be categorized for a particular product or brand.

Examples of instructions include:

Categorize tickets based on issue type.
Analyze customer sentiment.
Assign an appropriate priority level.
Return the response in JSON format.
Recommend escalation when necessary.

Well-defined instructions improve response quality, consistency, and reliability.

Apply the Agent’s Response to Ticket Fields

When enabled, the ticket fields selected below are filled automatically based on the AI agent’s response. Make sure the agent instruction mentions each field you select. By default, the AI agent’s response is added to the ticket as a private note. Enabling Update Ticket Fields Using AI Responses maps the response to the selected ticket properties instead, and no private note is added.

To automatically update ticket properties, check the box Apply the Agent’s Response to Ticket Fields. When enabled, the AI response is applied to the configured ticket actions instead of being added as a private note.

To configure this response mapping:

Check the Apply the Agent’s Response to Ticket Fields checkbox.
Choose the ticket action from the dropdown. Values for mapped actions are automatically applied to the corresponding ticket properties. Only supported values will be applied.
Click Add new action if more actions need to be mapped.
Once saved, the rule will run when a ticket matches the configured condition.

Response mapping works best when the AI Agent returns structured output, such as JSON. AI orchestration does not support nested JSON structures. As a result, nested fields (for example, status and reason) may not be processed and could be ignored. To ensure all data is captured and handled correctly, please provide values in a flat JSON structure whenever possible. Explore How to Update Ticket Fields Using AI Trigger Orchestration in BoldDesk.
If you use AI Trigger Orchestration to add tags, the tags must already exist and have been created in the portal; otherwise, they will be ignored.

How AI Trigger Orchestration Works

The following workflow occurs when a ticket matches a configured rule:

A ticket event occurs.
BoldDesk evaluates the configured rule conditions.
When the conditions match, the selected AI Agent is invoked.
The selected ticket context fields are sent to the AI Agent.
The AI Agent processes the information according to the configured instructions.
The AI Agent returns a response.
BoldDesk either:
- Adds the response as a private note, or
- Maps the response to configured ticket actions.

Use Cases

Automatic Ticket Categorization

Automatically classify incoming tickets into categories such as:

Billing
Technical Support
Sales
Feature Requests

Benefits:

Reduces manual classification
Improves routing accuracy
Speeds up triage

Automatic Priority Assignment

Analyze ticket details and assign an appropriate priority level.
Benefits:

Consistent prioritization
Faster incident response
Improved workload management

Sentiment Analysis and Escalation

Detect negative customer sentiment and escalate tickets when necessary.
Benefits:

Improves customer retention
Enables proactive support
Identifies high-risk conversations

SLA Risk Detection

Identify tickets that may breach SLA targets.
Benefits:

Improves SLA compliance
Supports proactive intervention
Reduces missed deadlines

Permissions

To create, edit, or manage AI Trigger Orchestration rules, your role must include the Manage Settings permission.

If you cannot access the AI Trigger Orchestration page, verify that the Manage Settings permission is enabled for your role. For more information about configuring role permissions, explore Managing Roles and Permissions in BoldDesk.

Administrators typically have this permission by default. For custom roles, an administrator must grant the Manage Settings permission before users can access AI Trigger Orchestration.

Troubleshooting

The AI Agent is not available in the dropdown

Verify the AI Agent has been created and published.

The rule does not run

Review the configured condition and confirm the ticket matches the rule criteria.

The ticket property is not updated

Confirm Apply the Agent’s Response to Ticket Fields (also referenced as “Map AI Response to Ticket Actions”) is enabled.
Confirm the selected action supports the returned value.
If the action is Add Tags, confirm the tag already exists in the portal.
If the AI Agent returns structured output, confirm the output is not nested JSON, because nested fields may be ignored.

The AI response is added as a private note

This is expected when response mapping is not enabled.

The AI output is not accurate

Review the selected Ticket Context Fields and remove unrelated context.
Update Additional Instructions to make the AI Agent’s task clearer and more specific.

Frequently Asked Questions

Why are AI credits required to use AI Trigger Orchestration?
AI credits are required because the feature invokes an AI Agent to analyse ticket information and generate a response. Each AI action may consume credits based on the AI usage in your BoldDesk account.
What happens if there are no available AI credits?
If there are no available AI credits, the AI Agent will not run when the orchestration rule is triggered. Make sure your account has sufficient AI credits before using this feature in an active workflow.
Can I use AI Trigger Orchestration without mapping the AI response to ticket actions?
Yes. If Map AI Response to Ticket Actions is not enabled, the AI Agent response is added to the ticket as a private note by default.
When should I enable Map AI Response to Ticket Actions?
Enable Map AI Response to Ticket Actions when you want the AI Agent response to update ticket properties automatically instead of adding the response as a private note.
Can I select multiple ticket context fields?
Yes. You can select one or more ticket context fields. The selected fields help the AI Agent understand the ticket details and generate a more accurate response.

Was this article useful?

Dislike

Comments (0)