The IQ series: three working apps, one fictional lender
Show the full caseHide the full case
- GovernIQ: decide which controls apply to the AI system, why, and who owns each one. Open GovernIQ (opens in a new tab) ↗
- SupplierIQ: check the supplier behind the model, and record what is still unanswered. Open SupplierIQ (opens in a new tab) ↗
- ImpactIQ: assess what the system could do to customers, and the evidence behind each claim. Open ImpactIQ (opens in a new tab) ↗
Why it is useful
- It turns an AI governance standard into a record a board can read in minutes.
- Missing evidence and unfinished decisions come first, not hidden inside a score.
- Every open item ends with a named owner and a next action.
What it does not do
- The software never makes the judgement. People write the decisions; the apps organise them and flag what is missing.
- The companies and data are invented. These are educational demos, not legal advice or a compliance check.
- The three apps are separate and do not share data.
Try it: Follow the 15-minute walkthrough → Or read the one-minute guide, or see how they were built and tested.
First look: what each one is for
The story, in five steps
-
ScopeGovernIQLive fictional demo
Northbridge starts by recording what its credit-decision model covers, who owns it and which controls apply, each with a reason. Fictional example.

Open full-size image ↗ -
The vendorSupplierIQLive fictional demo
Meridian’s model makes the decisions, so Northbridge checks the supplier before relying on it. The completeness measure describes the diligence record, not the supplier. Fictional example.

Open full-size image ↗ -
The impactImpactIQLive fictional demo
One human-written assessment of the credit model’s effect on customers, showing what still needs a decision.

Open full-size image ↗ -
The harmsImpactIQLive fictional demo
The matrix plots only the harms people recorded. It does not decide the risk.

Open full-size image ↗ -
What is still openGovernIQLive fictional demo
The record ends with open work grouped by owner, unassigned items first. Nothing is scored.

Open full-size image ↗
All three are live fictional demos. They are separate apps with no shared data, accounts or integration, and none of them makes a compliance determination.


