Evaluate support-ticket triage on three examples
Run a small ticket classification evaluation with fictional inputs, a saved model response, and explicit checks before building an integration.
What this exercise produces
This is an evaluation of classification suggestions for historical fictional tickets. It does not create a queue or route current incidents. The downloadable JSON contains the entire instruction and input. Keep the English fixture unchanged for the first run so the checks below match.
Run the fictional example
Install Node.js 22 or later on your computer. Download the request helper and input file below into a new private folder. In the Model Catalog, choose a text model with chat-completions support and create a model access key restricted to that model. Copy its exact model ID.
Create a file named .env in that folder using a text editor. Put DIGITALOCEAN_TOKEN= followed by your key on the first line, and DIGITALOCEAN_MODEL= followed by the model ID on the second. Keep this folder outside a repository and do not share .env. Open a terminal in the folder and run the command below. Each run uses billed model tokens.
Open the result JSON in your editor and read its draft field. It is a proposal awaiting review. HTTP errors, empty answers and truncated responses exit unsuccessfully. Use a new output filename for another run. If JSON was requested, parse the draft separately and reject missing fields, duplicate or unknown IDs, unsupported claims and invalid categories. Keep failures in the manual review list.
node --env-file=.env request.mjs triage.json triage-result.json
Download request.mjs · Download triage.json · Node.js · Model access keys
Compare the result with these checks
Expect exactly T-01, T-02 and T-03. T-01 is how_to with normal urgency and no escalation. T-02 is billing with a payment-dispute escalation. T-03 is access with urgent security escalation. Check the summaries preserve the issue without adding account facts.
A missing or invalid item goes to manual review, never to a default low-priority queue. These three checks demonstrate the format; they do not measure accuracy on your backlog. Evaluate a representative, independently labeled sample before implementing routing.