Triage Tickets: Claude + Make

A solo SaaS founder does his own support, and launch week buries him: password resets, billing confusion, and one quietly furious enterprise trial all land in the same undifferentiated queue. The urgent ones wait behind the trivial ones, because sorting them by hand is the work he never gets to before the next twenty arrive.

This build sorts them as they land. When a ticket comes in, Claude reads it and tags the topic, the priority, and the customer’s mood, right on the ticket in the help desk. The founder opens a queue that is already triaged, and works the hot ones first.

We wire it in Make, the first build in our customer support set. We use Zendesk as the help desk, but the shape swaps cleanly for Freshdesk, Help Scout, or Intercom. Claude only tags here; it does not reply, so there is no risk to the customer.

What does this workflow actually do?

In one line: a new ticket becomes a tagged ticket. Make watches your help desk. When a ticket arrives, Claude reads it and returns a topic, a priority, and a sentiment, and Make writes those onto the ticket as tags or fields. Your queue sorts itself; nothing is sent to the customer.

A few real cases, none of them the usual ones:

• The solo SaaS founder above, working the furious enterprise trial before the password resets.
• A camera-gear rental shop separating gear questions from damage claims automatically.
• An online course creator with a flood of login and access tickets at every launch.
• A regional internet provider splitting outage reports from billing questions on sight.

Reading a ticket is quick. Reading every one to sort the queue, all day, is the grind that buries the urgent ones. That sorting is exactly what Claude does in a moment.

Why use Make if Claude can already read and write?

Fair question. Claude is good at reading a ticket and judging what it is about and how urgent it sounds. It is not built to watch the help desk and write tags back onto each ticket. That is plumbing, and Make does plumbing.

Make handles the no-judgment steps: catching the ticket, writing the tags, running on its own. Claude handles the one step that needs reading, the triage call. Make is the mailroom; Claude is the one who reads each envelope and sorts it. Because it only tags, never replies, the customer never sees the automation, which makes it a safe first support build.

What do you need before you start?

• A Make account. The free 1,000 operations a month covers a busy queue.
• A Claude API key from the Claude console, not the chat app. New to Claude? See how to use Claude.
• A help desk with a new-ticket trigger and a way to write tags or fields. We use Zendesk.
• A short list of your topics and what each priority level means.

Two Make words, once. A scenario is one automation, the whole ticket-to-tag chain. A module is one box in it. Three boxes, one scenario.

How does the workflow work, step by step?

Three modules, left to right:

Module	App	What it does
1. Trigger	Zendesk	Fires when a new support ticket arrives
2. Brain	Claude	Reads the ticket and decides topic, priority, sentiment
3. Output	Zendesk	Writes the tags or fields back onto the ticket

Step 1: Watch your help desk

Create a scenario and add Zendesk, Watch Tickets. Connect your account and choose to watch new tickets. This trigger hands the next step the ticket’s subject and body. Using another help desk? Swap this one module for its new-ticket trigger.

Step 2: Let Claude triage it

Add a Claude module (Anthropic), Create a Prompt, paste your key, pick Sonnet. Give it your fixed lists:

Triage this support ticket. Reply on one line, pipe-separated.
Ticket: {{subject and body from Step 1}}

Topic (billing, bug, how-to, account, other) | Priority (low, normal,
high = money, outage, or an angry customer) | Sentiment
(calm, frustrated, angry). Decide only from the ticket text.

Defining the topics and what makes something high priority is what keeps the triage consistent enough to trust.

Step 3: Tag the ticket

Add Zendesk, Update a Ticket, point it at the ticket from Step 1, and map Claude’s topic, priority, and sentiment into tags or custom fields. Now your queue can be sorted and filtered by what actually matters.

How do you run it and check the result?

Click Run once on a test ticket. When I tested this, the triage matched my own read once I spelled out what made a ticket high priority, which was the one thing I tuned. Adjust the lists, then turn the scenario on so tickets tag themselves as they arrive.

After that, your queue stops being a flat pile. You filter to high-priority and angry first, and the trivial tickets wait, which is the whole point of triage and the thing solo and small teams never have time to do by hand.

What does this cost to run?

Piece	Free tier	If you outgrow it
Make	1,000 operations/month free	Core plan from about $9/month
Claude API	Pay per use	Triaging a ticket on Sonnet costs a fraction of a cent
Zendesk	Your existing plan	No extra cost
Your time	Seconds saved per ticket	Adds up fast at volume

Each ticket is three Make operations and a sliver of Claude. Even a heavy queue sits inside the free plan. For tiers, see our Make beginner’s guide.

What can go wrong, and how do you avoid it?

• Everything tags high priority. Define what high means with real criteria; tune it in testing.
• Topics drift. Give Claude a fixed list and an “other” bucket so the tags stay countable.
• It writes to the wrong field. Map Step 3 to the ticket from Step 1, and to real tag or field names.
• Customer data is sensitive. Send only the ticket text Claude needs, and keep help-desk access locked down.

Want more of this pattern? Our Make AI scenarios roundup has the next ideas.

How do you build this in Zapier or n8n instead?

The three jobs do not change, only the names do.

Job	Make	Zapier	n8n
Catch a ticket	Watch Tickets	New Ticket trigger	Zendesk Trigger node
Triage it	Claude module	Claude (Anthropic) action	Anthropic node
Tag the ticket	Update a Ticket	Update Ticket action	Zendesk node

Make and Zapier are the gentlest for a first build. Zapier vs Make vs n8n compares all three.

Common questions

Does the customer see any of this?

No. This build only tags the ticket internally. Nothing is sent to the customer, which makes it a safe place to start with support automation.

Do I have to use Zendesk?

No. Freshdesk, Help Scout, Intercom, and others have new-ticket triggers and tag actions. Swap the first and last modules; the Claude step is the same.

How does it decide priority?

By the rules you write into the prompt. Spell out what high, normal, and low mean for your business, and it follows them.

Can it route tickets to a team too?

Yes. Add a field for team and map it in Step 3, or add a router. Start with tags, then build on them.

Is the Claude chat app the same as the API?

Same models, different door. Make talks to the API, so you need a key from the console.

Sources and docs

• Creating a scenario (Make Help)
• Claude API overview (Anthropic)
• Zendesk Help Center
• Claude on Make (integrations)
• Help desk software — Grokipedia

Last reviewed: May 2026. Make, Claude, and Zendesk update their interfaces; check the official pages for exact button names.

You may also like

• AI Automation for Beginners (the hub)
• Triage email into Slack with Claude
• Draft Gmail replies with Claude + Make
• Make.com Complete Beginner’s Guide
• How to use Claude AI
• Zapier vs Make vs n8n: which to start with
• The AI Tools Directory