Every situation enters. Only some require action.
Within a deployment boundary, every available employer-paid healthcare situation enters The Engine from the beginning. Nobody selects which ones look suspicious first. The Engine does not run the underlying care — it is the system of record for whether responsibility was fulfilled, and it produces an answer only where the evidence supports one.
Every available employer-paid healthcare situation inside the deployment boundary enters The Engine at inception — not the ones someone has already flagged.
A situation enters as structured inputs — who is involved, what was owed, what was done, and the evidence for it. The Engine organizes supplied information; it does not claim to interpret completely unstructured material on its own, and it does not claim every source system is connected. What matters is that situations are not pre-filtered by whether they already look expensive or problematic. At the moment a situation enters, nobody knows which ones will require attention.
Every situation is processed by the same rules. The Engine organizes each one into the elements accountability actually depends on:
- Facts
- The observed context of the situation, preserved as recorded rather than overwritten.
- Participants
- The people and organizations involved, and the capacity each one holds.
- Obligations
- What was owed, and the moment a specific participant accepted a specific obligation.
- Authority
- Who is allowed to see, decide, or act — determined situation by situation, not just by who is logged in.
- Evidence
- The artifacts that demonstrate what happened, kept authoritative and immutable.
- Verification
- Whether the evidence has been independently confirmed, and to what degree.
- Attribution
- Which participant a result attaches to, and on what basis.
- Outcome
- What the record establishes — and, just as clearly, what it does not.
Most situations that enter will not require action. That is a valid, complete outcome — not a gap.
Because every situation enters, most of them resolve with the record simply showing that no action is currently required, or that the situation is being monitored because the present evidence is sufficient. A bounded answer of “no action shown” is a real answer. It means the record was examined under the same rules and found to support nothing further — not that the situation was ignored, and not that a problem is hidden.
The value of universal application is the denominator it creates: because everything entered, the record can distinguish the situations that genuinely require attention from the ones that do not, instead of only ever seeing the situations someone already noticed.
When a situation develops in a way the record actually supports, The Engine produces an answer. The answer names what the record establishes, who must act, what authority they hold, what evidence would close the open obligation, and what remains unknown. It does not manufacture a conclusion the evidence does not carry, and it stops where the evidence stops.
- New situation. Same structure.
- New facts and participants. Same rules.
- Each situation gets its own answer, built the same way every time.
- The conclusion is derived from the record, never authored around it.
Beneath the plain-language model, every Engine situation that requires action follows this exact ten-step sequence. The sequence is constitutional — it cannot be altered per client, per situation, or per participant.
- 1A situation enters.
- 2The governing obligation is identified.
- 3Responsibility and authority are assigned.
- 4The responsible party accepts or disputes.
- 5Execution is claimed.
- 6Evidence is attached.
- 7The result is verified, contradicted, or left insufficient.
- 8Delay remains visible while the obligation is open.
- 9Escalation occurs when the responsible action is not completed.
- 10The economic consequence remains connected to the record.
Once the shared record exists, the same record can be viewed by a physician, an employer, an administrator or vendor, an adviser, a fiduciary or legal reviewer, and a patient. These are not six separate products — they are authorized perspectives over one shared accountable record.
- The facts remain stable across every perspective.
- Authority and perspective determine what can be seen.
- The controlling question changes with the perspective.
- The projection and its boundary may change.
- The underlying record does not change.
Switching perspective does not grant authority. What you are allowed to do is checked every time, for that specific situation.
What does the system actually produce?
Continue to The Engine Answer