Datasets

Versioned input and expected-output pairs that power your evaluation suites. Select a dataset to preview rows or append new ones from CSV.

All datasets

5 datasets · 4,948 rows

support-tickets-q3

Labelled support tickets sampled from Q3 production traffic, PII redacted.

v7
Rows
2,400
Source
Production traces
Size
3.8 MB
Updated
2h ago · Mira
Used byTicket routing accuracy

Preview

First 6 of 2,400 rows

RowInputExpected outputTags
row_0001I was charged twice for the Team plan in October. Can you refund one?billing · high · refund duplicate chargebillingrefund
row_0002SSO login loops back to the sign-in page for everyone in our org.outage · urgent · escalate to on-callauthoutage
row_0003How do I export all my workspace data before we cancel?account · normal · explain exportaccountchurn-risk
row_0004Your API returns 500 on /v2/batches above 500 items.bug · high · request ID + engineering ticketapibug
row_0005Can we get invoices in EUR from next month?billing · normal · confirm currency changebilling
row_0006The mobile app logs me out every time I switch workspaces.bug · normal · collect app versionmobilebug

Upload CSV

Append rows to support-tickets-q3. Columns: input, expected_output, tags (comma separated).

A new dataset version is created; existing suites keep their pinned version.