Cross-session AI use case

[wabernathy]

I want to train the Atlassian AI to summarize certain Jira issues (bugs, stories) for release notes. This task does not fit naturally into Jira's AI's anticipated summary use case. I can badger and cajole the AI to produce what I want, but the slate is wiped clean when the job is done, and my AI training efforts amount to nothing.

I want to train the AI to summarize the underlying issue (not the ticket as a document) and to maintain a house style from session to session. I would like to associate this identity with one custom field ("Release notes description"), such that I (and my peers) can ask the AI to summarize the ticket, give a useful result, and learn and remember my company's house style from one session to the next. 

Is this feasible? 

[Jim Knepley - ReleaseTEAM · Accepted solution]

Hi @wabernathy 

There's a Rovo agent that looks like it fits your interests.

"Release Notes Drafter

Create comprehensive and clear release notes from a set of up to 20 Jira issues at a time with Release Notes Drafter. Summarize issues and group them into themes to save you time working out the best way to frame release notes for your work items."

The prompt can be adapted to support more than 20 issues if that's a concern. The prompt is too long to paste here, here's a link: https://pastebin.com/YUqbbuYS