Filters read one message at a time. Harm builds across the arc of a conversation. Step through one and watch the difference appear.
Three of the sixteen crisis categories in the standard scenario set. Each one is a situation a real person could walk into on a Tuesday.
Every reply below clears a content filter. Nothing in it is prohibited. Watch the eight dimensions move as the conversation goes on.
Bars show movement across the conversation, not published values. The composite is a weighted aggregate, not an average. The eight dimensions are weighted. The weighting is proprietary.
The score of the conversation so far, after each turn. A one-time reading finds the Gate failure. Monitoring is watching the arc, so the record shows movement well before that.
| Turn | As it ran | To the standard | Gate | Monitoring |
|---|
Looking back once. The Gate failure is on the record, traced to its turn, with the transcript attached. Everything before it passed.
Going forward, continuously. The same standard, run on live traffic, so sustained movement is on the record while it is still movement. Monitoring measures and reports. Nothing we run sits in your response path.
| Band | Range |
|---|---|
| Meets the standard | 85 to 100 |
| Approaching the standard | 70 to 84 |
| Below the standard | 50 to 69 |
| Far below the standard | 0 to 49 |
The bands describe where a measurement falls against the standard, not a verdict on the organization that built the system. Ikwe reports the measurement. What to do about it belongs to the operator.
Everything on this page is illustrative. The conversations are composites written for this walkthrough. The scores and the dimension movements are invented and belong to no real system. No model is named, scored, or compared here.
Monitoring measures. It does not intervene. Nothing we run sits in your response path, so nothing we run can add latency or break in production. A detection point is a point on the record, not a hand on the conversation.
This is not a report. An Ikwe report is evidence, not a certificate; no method available today is a certificate of compliance, and whether it satisfies a particular obligation is a determination for your counsel.
We measure. We do not grade, certify or recommend. We describe the standard. You decide what to change. We do not fix what we score, because an assessor who designs the fix is grading their own work.
The instrument is protected. The eight dimensions and the four bands are published. The scoring anchors, the dimension weights and the scenario set are not, because a public scenario set can be trained against, which would void every score ever produced with it.
Ikwe does not score whether a response was acceptable. Ikwe scores whether it was acceptable given what the system had already been told. That is why the unit of measurement is the conversation. See how the instrument works.
An Assessment runs your system against the same standard, over the same kind of arc, and comes back with every failure traced to the turn it happened on.
You do not have an answer on paper and someone has started asking. An Assessment tells you where you stand.
A board, a carrier or a regulator has a named question. An Audit is looking back once, scoped to it.
You cannot predict the shape of the question yet. Monitoring is the same standard, run continuously.
Thirty minutes. We will tell you which of the three fits, scope it in writing, and send pricing. Nothing connects until you sign off, and the record that comes back is yours. If none of the three is right for you yet, we will say so.