As a Knowledge Base Manager, I don't need to see every conversation users have with Rovo; I need to know whether Rovo's answers are helping my team, and when they're not, whether it's because content is missing from our knowledge base.
We don't need to read full conversations; that would be invasive and unnecessary. We just need aggregated quality signals that help us improve our documentation.
The request is to have a Rovo Quality & Knowledge Gap Dashboard for admins with:
Knowledge gap detection: When Rovo can't find relevant content to answer a question, flag the topic or category (not the full conversation) so KB managers know what documentation is missing
Topic clustering: Group unanswered or low-rated queries by theme (e.g., "billing process," "refund policy," "onboarding steps") so we can prioritize which articles to create or improve
Trend view: See if satisfaction is improving over time as we add content, confirming our KB efforts are paying off
We rely on team members manually submitting Jira tickets when Rovo gives a bad answer. But most users don't bother; they just move on. The knowledge gap stays invisible.
The core ask is simple: let us see where our knowledge base is failing our team, so we can fix it. Rovo already knows when it can't find an answer; we just need that insight surfaced to the people who can act on it.
