Matrix42 can also provide predefined AI-driven automation scenarios based on AI Functions. 

Depending on the installed product version, subscription, and activated extensions, these may include scenarios such as: 

Neo Agent for Ticket Triage (Ticket Preparation)

Analyzes incoming tickets and prepares suggested values before an agent starts processing them. 

This may include suggestions for: 

• Ticket type
• Category
• Service
• Impact
• Urgency
• Possible resolution 

It can:

• Determine whether the incoming ticket represents an Incident or Service Request
• Detect the appropriate category
• Determine impact and urgency
• Calculate priority
• Transform the ticket to the appropriate target type
• Apply required default values
• Route the ticket to the appropriate team
• Add journal entries for automated updates
• Notify relevant stakeholders

The goal is to reduce manual triage work and improve consistency in classification and assignment.

A typical workflow can start when a new ticket is created from sources such as email or the Self Service Portal.

Matrix42 Neo then analyzes the ticket and applies the configured processing logic before routing it to the responsible team.

Neo Agent for Ticket Service Detection

When a ticket is created without a service assigned, Neo analyzes the ticket content to detect the correct service automatically.

If the detection confidence is high, the correct service is assigned automatically. If confidence is lower, a service suggestion is shown to the agent for review instead. All actions are recorded in the ticket journal, including the detected or proposed service and the confidence level, so agents can see how the decision was reached.

Administrators can configure the confidence threshold and whether the feature runs in automatic or advisory-only mode.

It can be used to:

• Detect the affected or requested service automatically
• Enrich the ticket with the corresponding service information
• Improve downstream classification
• Improve assignment and routing
• Provide additional context for subsequent AI and automation steps

By adding service context early in the ticket lifecycle, subsequent workflows and service-management decisions can operate on more complete information.

Detailed behaviour:

• Only runs if Reported Service isn't already set.
• When a Solution is found, and the confidence is equal or greater than the threshold, the Service field is set automatically, and a journal entry (Service Updated by AI) records the detected service and confidence level.
• When no solution is found (or Advisory Mode is enabled), the Service field is left unchanged; a journal entry (Service Proposed by AI) records the proposed service and confidence level, noting it requires manual confirmation.

Neo Agent for Ticket Auto-Resolution 

Analyzes eligible tickets and attempts to identify a suitable resolution based on the available context and configured knowledge sources. AI Agents are controlled through their respective configuration, workflow, and security settings. Customers should validate the proposed behavior and results before enabling AI-driven processing for productive scenarios.

Depending on the configured scenario, the agent can:

• Analyze an incoming ticket
• Determine whether the ticket is suitable for automated resolution
• Find or generate an appropriate solution
• Apply configured workflow conditions
• Communicate the result to the requester
• Complete supported resolution steps
• Update the ticket and record the automated processing

The objective is to reduce manual effort for high-volume and repetitive requests while keeping the process governed by the configured workflow.

Tickets that cannot be resolved automatically can continue through the normal Service Desk process.

Detailed behaviour:

• When a Solution is found, and the confidence is equal or greater than the threshold, the the Solution field is updated, an email is sent to the requester, the ticket status is set to Resolved, and a journal entry (Ticket Automatically Resolved by AI) records the matched solution and confidence level.
• When no solution is found (or Advisory Mode is enabled),  a journal entry (AI Auto Resolution Agent Processed Ticket without Solution found) is added; no changes are made, and the ticket continues through standard processing.

Foundation & Detailed flow description

All Neo Agents are based at same Workflow Designer based workflow “AI Agent for Ticket Preparation”. 

That automatically means possiblity to customize Neo Agent to your needs

Step by Step

1. Read settings and prepare the ticket. Neo load the Service Desk→ AI Agents configuration, which every later decision uses. If image analysis is enabled, it extracts text and descriptions from the ticket's attachments and images and saves them on the ticket. A "Security Agent" flag is switched off while preparation runs and switched back on at the end.
2. Neo Analyzes attached document (Image or PDF), or embedded images to ticket description, when activated, optimized additional information for future AI use and saves it to hidden field, that will be later used by AI functions for this ticket.
3. Classify the ticket type. If type detection is enabled, the workflow determines whether the ticket is an Incident or a Service Request. When the AI's confidence score meets the configured threshold, the ticket is converted to the detected type and a journal entry records the change. Below the threshold, the suggested type is only logged, and the ticket is not changed.
4. Look for a solution (if ticket resolution is enabled). The workflow searches for a solution and checks the confidence score.
   • If the score is high enough, the solution is saved on the ticket and logged. If auto-resolve is enabled, a solution exists, the score meets its threshold, and one more required condition is met, the Incident or Service Request is automatically resolved and closed, with the resolution logged in the journal. This is one of the workflow's end points.
   • If the score is too low, the workflow looks up related records, sends clarification questions to the requester by email, logs the request, and creates a task. The workflow then waits for that task to be resolved before returning to the classification step to re-analyze the ticket with the new information. This is the workflow's only loop.
5. Detect the service (if service detection is enabled). When the score is high enough and the setting is in automatic mode, the service is set and logged as "set by AI." Otherwise it is only logged as "suggested by AI," if suggestions are enabled.
6. Routing and role (if ticket routing is enabled). The responsible role is detected and either set and logged as "set by AI," or only logged as "suggested by AI."
7. Category, impact, urgency, and priority (if ticket preparation is enabled). The category is detected and, above the confidence threshold, applied and logged; otherwise only suggested. If the role was not already set during routing, it is updated based on the category. Impact and urgency are each detected the same way, applied or suggested depending on confidence. Priority is then calculated and saved, with the change logged.
8. Notify and finish. Depending on whether the record is an Incident, a Service Request, or still a plain Ticket, the workflow reloads the record, notifies stakeholders by email, notifies responsible users with a push notification, and adds a journal entry. If the auto-resolve conditions are met, the record is resolved; otherwise the Security Agent flag is switched back on.

Activation

All Neo Agents are based at same Workflow Designer based workflow “AI Agent for Ticket Preparation”. 

As prerequstire for actviating any of Neo Agents, it is required to Enable this workflow from Application Menu →Workflow Designer→Workflows by selecting AI Agent for Ticket Preparation and click action Enable.

After this you can proceed to Configuration of Neo Agent

Configuration and Detailed Behavior for AI Agents

In the Service Desk Global Settings, an AI Agents tab controls three agents, each with its own enable toggle and Advisory Mode toggle, plus one shared Confidence Threshold:

Each agent has its own Advisory Mode (AI Suggestions Only) toggle (disabled by default). When Advisory Mode is enabled for an agent, that agent only adds its suggestion to the Ticket Journal. It does not apply the change automatically, regardless of confidence.

By default, the Confidence Threshold is set to 75%, shared across agents. This defines the minimum confidence level required for an AI Agent to automatically apply a change. Below this threshold, the system only proposes the change for manual review instead of applying it.

Image Analysis

A separate Image Analysis setting controls whether screenshots and images attached to a ticket are automatically analyzed when an agent opens the ticket. The generated description is displayed in the ticket's Image Analysis area and persisted, so reopening the ticket does not trigger another AI execution.

• Image analysis can be enabled or disabled at the tenant level. When disabled, no automatic analysis is triggered, and the Image Analysis area is hidden or shown as unavailable.
• It can also be enabled or disabled per queue. Disabling it for a specific queue does not affect other queues.
• Agents can trigger analysis manually from the ticket description or a journal entry, even when automatic analysis is disabled, where allowed by configuration.
• Up to 5 supported images per ticket are analyzed; additional images are skipped with a clear indication to the agent.

Clarification Feedback Loop

As part of AI Agent for Ticket Preparation, when the AI is not confident enough in a detected ticket type to proceed with automated solution discovery, the workflow can ask the requester clarification questions instead of stopping or escalating immediately.

• The requester is emailed clarification questions, and a task is created to track the response.
• The workflow waits for that task to be resolved, then re-runs ticket classification with the new information.
• This loop is configured through the AI Agent for Ticket Preparation workflow itself in Workflow Designer, so qualification thresholds and the clarification path can be adjusted to match an organization's requirements.

Journal Entries

All AI-driven journal entries are marked with a dedicated journal type so they're distinguishable from system or human entries. Administrators/agents can review the applied solution, confidence level, and automated steps taken directly in the ticket journal at any time.

Related Documentation 

For details on configuring AI service providers, managing AI Functions, supported deployment models, prerequisites, and troubleshooting, refer to: 

AI Functions & AI Service Providers: Configuration, Audit & Troubleshooting 

This article also provides an overview of predefined AI Functions, their configuration, and execution auditing.