Ticket Triage assessment

PERMISSIONS   Security level with Admin Feature Access permission to Cooper Copilot.

NAVIGATION   Left Navigation Menu > Admin > Admin Categories > Automation > Cooper Copilot

BEFORE YOU BEGIN  This feature may be hidden in your Autotask instance because it is not activated. If so, you can activate it on the Left Navigation Menu > Admin > Admin Categories > Activations page. Refer to Activations.

IMPORTANT  Digital Specialist for Ticket Triage will be released in phased groups beginning in May 2026, shortly after Kaseya Connect Las Vegas, for partners with Ultimate, ITSM+, Kaseya365 Ops, or Kaseya365 ITSM licensing.

IMPORTANT  To qualify for Early Access, partners must attend a mandatory training webinar covering data maturity, database preparation, and initial setup requirements. Following Kaseya Connect Las Vegas, eligible partners will be notified by email in phased groups and invited to attend a webinar. Early Access will be enabled no sooner than one week after webinar completion.

Overview

Manual ticket triage is a major bottleneck for modern MSP service desks. As ticket volumes and customer expectations increase, human-driven triage creates:

  • Delays in response and resolution
  • Inconsistent decisions across technicians
  • Higher operational costs

Ticket Triage addresses this by using AI to infer key ticket fields directly from ticket content and historical Autotask data, including:

  • Priority
  • Issue Type
  • Sub-Issue Type
  • Queue
  • Primary Resource

This enables faster, more consistent, and more scalable ticket management. The quality of these AI-driven decisions, however, depends on the quality of your Autotask data.

When fields are clearly defined and used consistently, Ticket Triage can make accurate recommendations and safely automate decisions. When data is vague, incomplete, or inconsistent, AI inferences become less reliable and automation confidence decreases.

A Data Maturity Assessment helps you:

  • Understand how ready your Autotask data is for AI-driven triage
  • Identify specific gaps in configuration and data quality
  • Prioritize improvements that will have the greatest impact on Ticket Triage accuracy

The more mature your data, the more value you will see from Ticket Triage, including:

  • Up to 50% reduction in manual triage effort.
  • Up to 95% improvement in consistency and standardization.
  • Up to 30% reduction in overhead labor costs linked to triage.
  • Up to 80% improvement in technician utilization and load balancing.

Ticket Triage is part of Kaseya's Agentic AI platform, housed in KaseyaOne. Commercials and pricing are being finalized, and UI changes may occur as those commercials are finalized.

To prepare Autotask environments for Ticket Triage and to improve results after initial deployment, MSPs must complete this assessment. Failure to do so may result in inaccurate outcomes, as insufficient contextual data will negatively impact inference accuracy. Refer to Part 1 - Prepare your data for Ticket Triage of the quick start guide below for more information.

Who should complete it

  • Service desk managers
  • PSA administrators
  • Operations leaders responsible for process and data quality

What an assessment covers

You will need to review and rate your data maturity in the following areas:

  • Priorities - How clearly you express business impact and urgency.
  • Issue and Sub-Issue Types - How well you describe problem types.
  • Queues - How clearly you define ownership and responsibilities.
  • Skills - How accurately your skills library reflects real capabilities.
  • Ticket Content - How consistently technicians capture clear information.
  • KPIs and Data Hygiene - How reliable your operational metrics and updates are.

How to work through the assessment

  1. Review each focus area
    Read the guidance for each area (Priority, Issue Types, Queues, Skills, Ticket Content, KPIs).
  2. Compare to your current configuration
    Look at how your Autotask system is configured today and how your team actually uses it.
  3. Identify gaps and risks
    Note where your fields are vague, inconsistent, or unused, and where processes are not followed.
  4. Use the checklists
    At the end of each section, use the readiness checklists to decide whether that area is:
    • Ready for Ticket Triage, or
    • Needs improvement before you rely on AI automation.
  5. Prioritize improvements
    Focus first on changes that:
    • Clarify priorities and queues
    • Reduce duplicates and vague entries in skills and issue types
    • Improve the quality of ticket descriptions and notes

How this connects to Ticket Triage

Ticket Triage learns from your historical tickets and current configurations. This ensures:

  • Your fields and descriptions accurately describe how work should be handled
  • Your tickets contain enough context for AI to recognize patterns
  • Your KPIs reflect real operational performance, not data noise

Once you complete an assessment and address the highest-priority gaps, you can enable Ticket Triage with greater confidence that its recommendations and automated decisions will align with your service standards.

Use this guide to prepare your Autotask data and enable Ticket Triage so that the AI can accurately prioritize, categorize, and assign tickets from day one.

What this quick start guide covers

  1. A brief description of how Ticket Triage works.
  2. A step-by-step guide to prepare your Autotask data.
  3. A setup guide to deploy Ticket Triage on release day.

How Ticket Triage works

Ticket Triage uses Kaseya's Agentic AI engine to extract and understand context from your tickets. It analyzes ticket content and historical Autotask data to automate:

  • Prioritization
  • Field categorization
  • Assignment to queues and resources

The AI engine interprets context from:

  • Existing ticket history (titles, descriptions, and notes)
  • Queue and priority descriptions
  • Skills and skill descriptions
  • Issue and Sub-Issue Types

Once data is analyzed, Ticket Triage can auto-apply or suggest values for these fields:

  • Priority
  • Issue Type
  • Sub-Issue Type
  • Queue
  • Primary Resource

Future releases will expand this field set to include additional ticket fields, including User Defined Fields (UDFs).

NOTE  Ticket Triage does not understand intent like a human. It recognizes patterns. Clear, consistent data creates clear patterns and better AI results.

Part 1 - Prepare your data for Ticket Triage

Complete the steps in this section before or during the Ticket Triage release to ensure your data is ready for accurate inference.

Step 1 - Add Priority descriptions (High Impact)

Add a clear, business-focused description for every active ticket priority. The AI uses these descriptions to understand business impact and urgency. Refer to Task & ticket priorities for more information.

Where to configure

  • Left Navigation Menu > Admin > Features and Settings > Service Desk (Tickets) > Task and Ticket Priorities

Recommended structure / best practices

  • Use the formula: Impact > Scope > Urgency > Business Consequence.

Example (Good): "For business impacting issues affecting multiple users or core business systems. Tickets under this priority require immediate service desk attention and response to prevent client operational disruptions."

Example to avoid: "High priority issues" | "Critical items from VIP"

Why this matters: The AI prioritization agent reads your priority descriptions as operational policy. Clear descriptions remove guesswork and align the AI with your business rules.

Step 2 - Add or refine queue descriptions (High Impact)

Ensure every active Queue has a description that clearly states what belongs in the queue and what does not. Refer to Service Desk queues for more information.

Where to configure

  • Left Navigation Menu > Admin > Service Desk (Tickets) > Queues

Recommended structure / best practices

  • Use the formula: Work Type > Inclusions > Explicit Exclusions.

Example (Good): "Handles Tier-1 end-user incidents related to workstation issues, login problems, and standard application access. Excludes server, network, and security incidents."

Example to avoid: "General Support"

Why this matters: The assignment agent compares queue descriptions with ticket content to find the best routing. Vague or generic queues lead directly to misrouted tickets.

Step 3 - Improve Issue and Sub-Issue type naming

Review all Issue and Sub-Issue types to ensure the names themselves clearly indicate the type of work or problem. Long descriptions are not required; clarity in naming is. Refer to Adding issue types and sub-issue types for more information.

Where to configure

  • Left Navigation Menu > Admin > Features and Settings > Service Desk (Tickets) > Issue and Sub-Issue Types

Recommended structure / best practices

  • Use clear, specific, outcome-oriented names.
  • Avoid ambiguous or duplicate meanings.
  • Ensure Sub-Issue types further refine the parent Issue Type.

Example (Good): Issue Type: "Email Delivery and Access" | Sub-Issue: "M365 Mailbox Authentication Failures"

Example to avoid: Issue Type: "Email" | Sub-Issue: "Problem"

Why this matters: The categorization agent uses Issue and Sub-Issue types as semantic anchors. Clear naming makes it easier for the AI to map ticket language to the correct problem category.

Step 4 - Review your Skills library (Required Dependency)

If you use Skills in Autotask, review and clean up your skills before enabling Ticket Triage. Refer to Managing certificates, degrees, and skills for more information.

If you do not currently use skills, or your skills are outdated and unused, you can skip this step for now. You will activate the new system skills on release day.

Where to configure

  • Left Navigation Menu > Admin > Features and Settings > Company Settings and Users > Resources/Users (HR) > Skills Management > Skills

Recommended structure / best practices

  • Identify and remove duplicate skills.
  • Add descriptions that clearly describe the technology and task capability.
  • Use the formula: Technology > Task Capability > Context.

Example (Good): "Troubleshooting Microsoft 365 mailbox access, authentication, and mail flow issues."

Example to avoid: "Email support"

Why this matters: The Skill Matching agent infers required skills from tickets and maps them to resources. Duplicate or vague skills introduce ambiguity and reduce assignment accuracy.

Part 2 - What to do on release day

On the day Ticket Triage is released to your environment, complete these steps to enable and control the automation.

Step 1 - Activate the new System Skills

System skills will be ACTIVE by default. In order to assign tickets based on skills, the AI engine needs a library of skills with descriptions that provide good context. To help you, Autotask includes a curated library of more than 240 system-provided skills, researched and aligned with ITIL, ITSM, and MSP best practices.

If you have your own set of custom skills, you may use them alongside system skills and you have the option to deactivate system skills in bulk, by category, or individually.

Where

  • Left Navigation Menu > Admin > Features and Settings > Company Settings and Users > Resources/Users (HR) > Skills Management > Skills

On release day, you can:

  • Review the system skills library.
  • Deactivate system skills individually.
  • Deactivate system skills by category.
  • Deactivate system skills in bulk if desired.

Recommended approach

  • If you have no existing skills, keep the system skills library active and use it as your starting point.
  • If you already have skills, review them first and remove or consolidate duplicates before deciding whether to deactivate any system skills.

What not to do

  • Avoid creating or keeping duplicate skills that overlap with system skills.
  • Avoid leaving vague skills without clear descriptions, as they reduce assignment accuracy.

Why this matters: The Skill Matching agent infers required skills from tickets and maps them to resources. A clean, well-structured skills library gives the AI confidence to make accurate assignment recommendations.

Step 2 - Configure Ticket Triage Settings

Enable Ticket Triage and define the data and conditions the AI will use.

Step 2.1 - Select historical tickets for analysis

Where

  • Left Navigation Menu > Admin > Automation > Cooper Copilot > Cooper Copilot Settings > Ticket Triage > Ticket Data Analysis

On release day, you can:

  • Select ticket statuses and queues for AI analysis.
  • Limit analysis to completed or stable tickets such as Resolved or Waiting on Client.
  • Focus on queues that best represent your core service desk operations.

Recommended approach

  • Choose tickets that are unlikely to be changed further.
  • Favor tickets with meaningful descriptions and technician notes.
  • Exclude experimental, test, or low-quality data sets.

What not to do

  • Do not include tickets that are constantly updated or that lack meaningful descriptions and notes.

Why this matters: The AI engine learns from the language and patterns in your historical tickets. Clean, stable tickets with good notes and accurate fields produce stronger, more reliable recommendations.

Step 2.2 - Configure workflow rules with Ticket Triage actions

Where

  • Left Navigation Menu > Admin > Automation > Cooper Copilot > Cooper Copilot Settings > Ticket Triage > Triage Actions

On release day, you can:

  • Create new workflow rules to identify which tickets should be triaged.
  • Update existing rules to add Ticket Triage actions.
  • Set confidence thresholds and choose per-field actions.

Recommended approach

  1. Create or update workflow rules to define events, conditions, and timing for triage.
  2. In the Actions section, enable the Ticket Triage actions:
    1. Set a Confidence Threshold (default is 95; you can use any value from 1 to 100).
      • The confidence threshold controls when the AI will auto-apply versus suggest values for each field.
    2. Choose an action per field (Primary Resource, Queue, Priority, Issue Type, Sub-Issue Type):
      • None - no change for this field.
      • Suggest - the best match is suggested for a user to accept or decline via the Triage Dashboard or within the ticket.
      • Auto-apply - the best match is automatically applied when the confidence is at or above the threshold; otherwise, a suggestion is presented.

What not to do

  • Do not duplicate existing workflow rules that already manipulate any of the fields managed by Ticket Triage (Priority, Queue, Issue or Sub-Issue Type, Primary Resource).

Why this matters: Ticket Triage relies on workflow triggers to start automation. Using Autotask workflow rules lets you adopt AI-driven triage at your own pace while avoiding conflicting automation.

Step 3 - Select which queues and resources participate

Where

  • Left Navigation Menu > Admin > Automation > Cooper Copilot > Cooper Copilot Settings > Ticket Triage > Assignment Eligibility

On release day, you can:

  • Select the Queues that can receive assignments from Ticket Triage.
  • Select the Resources that can be assigned:
    • By Department
    • By Workgroup
    • By individual resource

Recommended approach

  • Limit eligibility to queues actively used by your service desk.
  • Include only resources who should receive AI-driven assignments.
  • For co-managed environments, group co-managed resources into a workgroup and explicitly decide whether to include them.

Co-managed environments

  • Group co-managed resources into a Workgroup.
  • Include that Workgroup if you want co-managed users to participate in the triage process.
  • If included, they become part of the main pool of resources the system can assign to.

What not to do

  • Do not include inactive queues or resources.
  • Do not include co-managed resources if you do not want them receiving AI-driven assignments.

Why this matters: Assignment Eligibility defines the safe play area for the AI. Limiting eligibility to real, active service desk queues and resources prevents misdirected assignments.

Part 3 - Writing high-quality ticket content

The quality of your ticket content directly affects Ticket Triage accuracy. Use a simple writing formula for all ticket descriptions and notes:

Recommended writing formula

  • Meaning > Business Intent > Desired Outcome > Exceptions.

Examples

Bad description

"Email not working."

Good description

"User cannot send outbound emails from Outlook after password reset. Error indicates authentication failure. Impacts customer communication. No workaround available."

TIP   Encourage technicians to write brief but complete descriptions. A few extra seconds per ticket can dramatically improve AI recommendation quality over time.

Next Steps and Readiness Checklist

Before release day

  • [ ] All active priorities have clear, business-oriented descriptions.
  • [ ] All key queues have clear descriptions with inclusions and exclusions.
  • [ ] Issue Types and Sub-Issue Types use clear, specific names.
  • [ ] Skills library is reviewed and duplicate skills are removed (if used).

On release day

  • [ ] System skills are reviewed and adjusted (kept active, or deactivated in bulk, by category, or individually as needed).
  • [ ] Historical tickets for analysis are selected.
  • [ ] Workflow Rules are configured with Ticket Triage actions.
  • [ ] Confidence thresholds and per-field actions are defined.
  • [ ] Eligible queues and resources are selected for assignment.

Once you complete these steps, your environment is ready for Ticket Triage to deliver faster, more consistent, and more scalable ticket handling with AI-driven recommendations and automation.