Automate support ticket reporting
A clear backlog digest in #support every morning — SLA breaches, aging tickets, and who's overloaded — before anyone opens Zendesk.
Reporting on the support backlog by hand means pulling open tickets, spotting SLA breaches, and writing it up — before stand-up, every morning.
Relay.app writes the report so no one hand-counts the queue at 9 AM: every weekday it pulls open tickets from Zendesk, Claude flags SLA breaches and aging buckets, the support lead gives a quick review, and the digest posts to #support before stand-up.
Here's how it works, step by step
Every weekday at 9 AM, the backlog report runs itself
- Every weekday morning at 9 AM in your team's timezone, the workflow fires on its own — before the support stand-up.
- No one has to remember to open Zendesk and hand-count the queue.
Pull every open ticket from Zendesk and flag SLA risks with AI
- Relay.app connects to Zendesk and pulls every open ticket — status, priority, assignee, last reply, and SLA timer.
- Claude sorts them into aging buckets, flags SLA breaches and tickets with no reply in 24h+, and tallies open load by assignee.
Draft the backlog digest, review it, and post to Slack
- Claude turns the analysis into a Slack-ready digest — open count and day-over-day change, SLA breaches, the oldest tickets, and load by assignee.
- Before it lands, Relay.app DMs the support lead for a quick edit or sanity check.
- Once approved, the digest posts to #support so the queue is clear before anyone touches it.
Append a copy for backlog trend tracking
- The same digest gets appended as a row in a running "Support Backlog" Google Sheet, so the support lead can chart open count, SLA breaches, and aging week over week.
- Over a few weeks it turns into a clear trend line — whether the backlog is growing, where SLA breaches cluster, and who's consistently overloaded.
Make it yours in Relay.app
Start from this template — describe what you want in plain English and Relay.app's copilot builds it for you, then swap any app, change the AI model, or tweak any step until it fits your team.