Support

Classify Support Tickets with AI

Billing, Technical, Account, How-To, Feature Request — ticket triage with definitions that decide the borderline cases for the model.

Overview

Ticket triage fails at the borders: is a "payment page crashes" ticket Billing or Technical? Without definitions, the model decides differently every run. This setup classifies tickets into five routing categories whose one-sentence definitions draw the borders deliberately — How-To Question is defined as "nothing is broken" precisely to keep it apart from Technical — and reports a high/medium/low confidence with each label, so your queue can auto-route the confident calls and surface the uncertain ones to a human.

Workflow

Generate and pipe tickets through
One ticket per call; the prompt reads the ticket text from the input below it.
Check the border definitions
"Nothing is broken" in How-To Question and "product defects" in Technical are doing the triage — adjust them to where YOUR borders run.
Route by confidence
high → straight to the queue; medium/low → human review. The confidence contract exists for exactly this split.

Why This Works

Definitions written to exclude each other settle the borderline tickets
Single-label edge rules force a primary-purpose decision instead of a coin flip
Self-reported confidence makes partial automation safe — automate the easy 80%

Best for

Support teams triaging by hand today
Queues where Billing-vs-Technical borderline tickets cause bouncing
Helpdesk automations with a classification step

Not for

Pulling ticket fields (product, severity, summary) — that's the Extraction Prompt Generator
Writing the reply — triage decides where, not what to answer

Use cases

Routing inbound tickets to billing, technical, and account queues
Keeping triage consistent across shifts, agents, and model runs
Auto-routing high-confidence labels and reviewing low-confidence ones

Customize This Resource

Opens this setup in Data Classification Prompt. Generate to get the full classification prompt — then adjust the labels, ambiguity mode, and confidence output.

Open in Data Classification Prompt

Prompt Template

TASK
Classify incoming support tickets into routing categories for the support queue.
Assign exactly one of the labels defined below to the text in the input.

LABELS
- Billing: Questions or problems about invoices, charges, refunds, or plan pricing.
- Technical: Product defects, errors, crashes, or broken integrations.
- Account: Access, sign-in, password, or account management matters.
- How-To Question: A request for guidance on using the product as designed — nothing is broken.
- Feature Request: A suggestion for a new capability or an improvement to an existing one.

CLASSIFICATION RULES
- Read the entire text before deciding — the deciding signal may come late.
- Judge by content and intent, not by tone, length, or formatting.
- Match the text against the definitions, not just the label names.
- Use only the labels defined above — never invent a new label.

EDGE CASE RULES
- If the text fits more than one label, choose the one matching its primary purpose — what the writer mainly wants.
- If two labels fit equally, prefer the more specific definition over the more general one.

AMBIGUITY POLICY
- Choose a label only when its definition genuinely covers the text.
- When torn between two labels, re-read both definitions and pick the closer one — do not guess from the names.

CONFIDENCE
- With each label, report a confidence level: high, medium, or low.
- high = the text unambiguously fits the definition; medium = a good fit with minor uncertainty; low = the best available choice among the labels.
- Confidence reflects your certainty about the fit — it is self-reported, not computed.

OUTPUT FORMAT
Return your classification as a single valid JSON object with a "label" key and a "confidence" key.

EXAMPLE OF A VALID RESPONSE
{
  "label": "Billing",
  "confidence": "high"
}

OUTPUT RULES
- Return only the JSON object — no text before or after it.
- Do not wrap the output in markdown code fences.
- Return the label text exactly as defined above — no paraphrasing, no abbreviations, no new labels.

Related Resources

Explore more resources that work well with this one.

Used in Playbooks

Complete workflows that include this resource as a step.

AI Customer Support Workflow Workflow · 4 steps

Tip: Save time by exploring related resources and tools that integrate with this workflow.

Explore all resources