Support Ticket Auto-Categorizer

Overview

This flow triggers on every new submission to a Microsoft Forms support intake form, runs the ticket description through an AI Builder text-classification model to predict the ticket's category, picks the matching Teams channel via a 5-way Switch (Software / Hardware / Network / Access Request / Other), creates a Dataverse row in the flowlibs_supportticket table, and posts a triaged ticket card into the routed support Teams channel.

Use Case

Help desks burn time triaging incoming tickets — reading the description, deciding the queue, manually pinging the right team. Misroutes are common and slow down resolution. This flow uses an AI Builder text-classification model trained on the org's historical tickets to predict the category and route the new ticket to the matching Teams channel automatically, persisting an audit trail in Dataverse.

The flow is ideal for teams that:

• AI Builder text classification keeps tickets out of the wrong queue.
• Forms-driven intake means submitters use a familiar UX (no portal).
• Dataverse persistence enables Power BI dashboards on volume + accuracy.
• One env var per category-channel mapping — re-route a queue without flow edits.
• Switch with a Default case guarantees nothing is lost when the model can't classify confidently.

Flow Architecture

Environment Variables

Connectors & Connections

Note — All connections are referenced as solution connection references; the flow is portable between environments as long as a connection is mapped at import time.

Customization Guide

Almost every realistic variant of this flow can be implemented by changing environment variable values. A few cases require small edits inside the flow definition — those are called out explicitly below.

Add a low-confidence escalation

Insert an If after the prediction that checks varPredictionConfidence < 0.7 and routes to a human-review channel before the main Switch.

Notify the submitter

Add an Outlook Send Email (V2) after the Teams post acknowledging receipt and including the routed channel name.

Add SLA fields

In Create Support Ticket Row, set due_date based on addDays(utcNow(), n) driven by category — Network 1 day, Software 3, etc.

Categorize via AOAI instead

Swap the AI Builder action for an Azure OpenAI HTTP call when there isn't enough labeled data for AI Builder.

Add an attachments path

Use Forms's Get attachment content to pull file uploads, then store them on the row's Notes annotation table.

Round-robin within a queue

After the Switch, query Dataverse for the queue's on-call user and @-mention them in the Teams post.

Escalate VIP submitters

Look up varSubmitterEmail in an Office 365 Users action and bypass the Switch for users in a specific group.

Key Expressions

The flow is intentionally light on Power Fx / WDL gymnastics — the heaviest expressions are the branch-name concatenation and the approval outcome check. They are listed below in the order they appear in the flow.

outputs('Get_Ticket_Response_Details')?['body/' + parameters('flowlibs_SupportTicketQuestionId')]

outputs('Get_Ticket_Response_Details')?['body/responder']

first(body('Classify_Ticket_Text_with_AI_Builder')?['predictionOutput']?['result'])?['name']

first(body('Classify_Ticket_Text_with_AI_Builder')?['predictionOutput']?['result'])?['confidenceScore']

variables('varPredictedCategory')