Freddy AI Summarize is genuinely useful, but generation is fully manual today, an agent has to open the ticket, click Ticket Summary, wait for Freddy, then click Save Summary. There's no auto-generation, no API to trigger it, and no webhook when one lands.
In practice this means summary coverage depends entirely on whether an agent remembered to click two buttons (even making a Summary required to Resolve or Close the ticket may help). Anything built on top of summaries, analytics, QA, post-incident reviews, knowledge mining, is working off a biased sample of tickets. There's also no way to backfill historical tickets or wire summary generation into automation.
For example:
- A setting to auto-generate the summary on ticket resolution or closure. Maybe a toggle in Freddy AI Copilot settings. By the time the ticket hits Resolved/Closed, the summary is already there. This solves coverage in one move.
- An API endpoint to trigger generation: e.g.
POST /api/v2/tickets/{id}/summary. Useful even if (1) ships, because it covers historical backfill and custom automation. - A webhook for summary creation: e.g.
ticket.freddy_summary_createdwithticket_id,summary,generated_at. Eliminates polling for anyone consuming summaries downstream.
Freddy Summarize is the one part of the ticket lifecycle that isn't yet exposed to webhooks or APIs. Closing that gap would multiply its value without changing the core feature.
Thanks for considering!
