Critical risks current within a week

Request: „Go through every red risk ticket and nudge the owners with concrete proposed actions.“

1. The AI reads all risk tickets through the ticket system's API
2. It rates each ticket against fixed criteria — assessment, control date, mitigation, contingency plan, recency, linkage — and proposes mitigations; a dashboard shows a traffic light and open points per ticket
3. It drafts a friendly comment per ticket: a call to action plus recommended measures for the owner — after approval the comments are posted in the ticket system
4. One week later it counts what has been updated or closed

Risk tickets need continuous upkeep — mitigation, due date, owner feedback; in any organisation that runs through regular review meetings.

Set the criteria, check the tone, approve the write run. Owners decide whether and how to mitigate; the AI only asks and proposes. No ticket is closed automatically.

The query and the officially reported figure differed by a few tickets — the authoritative data source first had to be agreed with risk management.

Any management team tracking risks in a ticket system. A comment with a concrete proposal works better than a review meeting and needs no extra appointment.

Risk management lives on upkeep. The agentic PMO reads all roughly 100 active risk tickets through the ticket system's API, rates them against fixed criteria and drafts a friendly comment per ticket with a concrete proposal, posted after approval: which mitigation fits, by when, and what the owner can add. That reached around 90 tickets. Within a week the owners had updated or closed a good half of them, and a good three quarters of the critical red risks — without an extra meeting. The reason for the speed: the comment lands where the owner already works, and it comes with a proposal. That lowers the bar to act immediately.