Watch knowledge health
Who can do this:Every role
Transcript
- Knowledge health shows how complete, verified and current your knowledge is, and how that changes over time.
- Items by status show how much is verified and how much still waits for review.
- Every count leads straight to the work: contradictions, readiness findings and incidents.
- Gaps are listed too, such as the questions the assistant could not answer.
- Expert dependency shows whether verification rests on too few people.
Knowledge health shows the state of your knowledge in numbers, and how it develops. The dashboard carries a short version of it; Details opens the full page.
What it shows
Section titled “What it shows”| Card | What it tells you |
|---|---|
| Trend over 90 days | The verified share, the items awaiting review, open contradictions and obsolete sources, day by day. |
| Knowledge items by status | How many items are AI extracted, unverified, confirmed, approved, disputed or closed. Each number opens the matching list. |
| Review and quality | What waits for people: review, contradictions, readiness findings, model mappings, incidents and negative feedback. Each number leads to the work. |
| Sources by status | Active, obsolete, superseded, unverified and rejected sources. |
| Models without documentation, models with few closed cases | Where knowledge is thin. |
| Unanswered questions | What technicians asked in the last 30 days that the knowledge could not answer, with a link to each conversation. |
| Expert dependency | How verification decisions are spread over the experts, so the knowledge does not rest on one person. |
Good to know
Section titled “Good to know”- Propagate obsolescence marks the knowledge from obsolete and superseded sources at once; it also runs by itself when a source changes.
- Unanswered questions are good starting points for a capture campaign.