Evidence record
What the source or system actually says, with provenance retained.
How RILayer works
From signal to governed, accountable action.
RILayer introduces structure at the point where information becomes consequential: when a person or team interprets what they have seen and decides what should happen next.
Governed decision path
Signal → evidence → review → accountable action.
The mechanism is designed to make uncertainty, escalation and human authority visible before consequential action proceeds.
Human-in-the-loop
Non-advisory
Reviewable evidence
Designed for transferability
What RILayer makes visible
Make the path from information to accountable action reviewable.
The purpose is not simply to produce another recommendation. It is to make the path from information to accountable action more visible and reviewable.
Uncertainty
What is missing, unresolved or still requires interpretation.
Escalation
What cannot safely or properly proceed without further human review.
Human decision record
What the authorised human decided, why, and at which consequential gate.
Change trail
Material corrections, holds, overrides, disagreements and changes retained for review.
The operating journey
RILayer slows the right moments, not every moment.
The objective is not bureaucracy. It is targeted structure where speed, uncertainty, pressure or AI reliance can otherwise hide weak reasoning and unclear responsibility.
- 01
Signal enters
A signal may be an AI output, workforce concern, risk indicator, customer issue, manager request, policy pressure, operational escalation or strategic opportunity. - 02
Interpretation is exposed
RILayer makes visible what is known, assumed, inferred, missing or being treated as certain without sufficient support. - 03
Governance controls apply
Evidence, Decision Readiness, Governance Risk, boundaries, routing and accountability are examined before the organisation treats the signal as an authorised route forward. - 04
A governed outcome is recorded
Authorised human judgement may lead to action, pause, escalation, further evidence, professional review or a decision not to proceed, with a reviewable record of what changed and why.
Discernment before action
Not everything that enters a decision environment is evidence.
RILayer helps organisations distinguish machine output, possibility, pressure and assumptions from the authorised evidence needed for accountable action.
Information is not action. AI output is not judgement.
The most important control point is what happens between receiving a signal and treating it as a justified organisational response.
Evidence
What supports the interpretation?Separate verified evidence from assumptions, AI output, pressure, confidence and reaction.
Decision Readiness
Is the next governed decision properly ready?Check whether sufficient authorised evidence, clarity, boundaries and authority exist for the next step. This is not a score of the person.
Governance Risk
What could go wrong in the decision process?Surface risk arising from interpretation, evidence, process, system input, boundaries, authority or routing — never a person-risk score.
Boundaries
What limits apply?Make professional, statutory, organisational and authority limits visible before they are crossed.
Route
What should happen next?Choose a proportionate governed route: proceed, pause, seek evidence, escalate or use authorised human judgement where the evidence supports it.
Accountability
Who owns the reasoning?Retain a clear human owner for interpretation, action, evidence and later review.
The six controls do not decide the outcome. They make the conditions for proportionate human judgement explicit before action proceeds.
Demonstrators
Follow the mechanism through a bounded case.
YT-WM-001 is the current reference proof-of-concept. It shows where machine output stops, where governance controls apply, where uncertainty remains visible and where accountable human judgement resumes.
Youth Transition · YT-WM-001
A machine says “do not progress.” RILayer makes the decision governable.
The fictional Chloe Taylor case begins with a simulated screening-system recommendation: DO NOT PROGRESS. The browser demonstrator exposes evidence, readiness, governance risk, boundaries and routing without manufacturing a replacement score or taking the final decision away from the accountable people.
01
Case context
02
Machine signal
03
Evidence check
04
Decision readiness
05
Governance risk
06
Boundaries
07
Governed route
08
Human decision
09
Decision record
Designed for adaptation
The governance mechanism may transfer. The evidence has to earn that claim.
The architecture is intended to transfer between decision environments, while sector expertise, professional responsibility and evidence requirements remain context-specific. Youth Transition remains one bounded controlled proof lane. Other applications remain adaptation and validation hypotheses until tested.