Every morning it checks a Jira project against agreed thresholds, forecasts each milestone from team velocity, and alerts me to anything new. Every Monday it writes the steering-committee status report. Total running cost: €0.
A demo hardware project (battery management system) with fictional data, updated automatically by the workflow below.
Loading live data…
As a technical project manager on hardware programs, I spent hours every week pulling Jira data together for C-level steering meetings. Risks like a supplier lead time or a lab that was already booked often became visible only when the report was being written. I wanted the project to report on itself.
How it worksProgress, RAG status, forecasts and alerts come from transparent rules. The AI (Gemini) only turns those facts into a readable report, and it must cite a Jira ticket for every risk. It can't make up a status.
Remaining story points ÷ average velocity of the last three sprints gives a forecast finish date per milestone. A milestone is "at risk" when that forecast passes its planned date, weeks before the date itself is missed.
Overdue work, blocked tickets, scope growth above 10% a week, a velocity drop above 20%, and a sprint more than 20% above capacity. Each is a plain-language setting that a project team can agree on and change.
Every alert has an identity, so it's sent once, when it first appears, not every morning until it's fixed. Jira's own automation handles the instant "ticket blocked" email. Each tool does the job it's best at.