Observed results
Measured in a real program with the population, period, comparison, and limitations attached. An observational result does not prove Rely caused the change.
Proof
A result starts with the record behind it.
We define the population, work performed, disposition, comparison, and limitation before a number is ready to publish.
Published results include the population, comparison, period, and limitations needed to interpret them.
See the fields behind every reported result.
The number travels with the fields needed to interpret it.
- 01Eligible population
- 02Exposure or intervention
- 03Outcome
- 04Denominator
- 05Method
- 06Limitations and permission
Observed association is not presented as causation.
See the fields behind every reported result. The number travels with the fields needed to interpret it. The diagram shows Eligible population, Exposure or intervention, Outcome, Denominator, Method, Limitations and permission. Observed association is not presented as causation.
The required anatomy
Eight fields before a result is publishable
A rate without its population, method, and limits is not evidence.
- 01
Eligible population
Who was in scope and the denominator used.
- 02
AI action
What the agent was permitted to do and where it stopped.
- 03
Navigator action
What a navigator personally worked, when included.
- 04
Work returned
What came back to the customer team.
- 05
Disposition
Completed, closed with a reason, or accepted by a named owner.
- 06
Source reconciliation
Where records disagree, the source system wins and the gap is reported.
- 07
Outcome
Measure, period, comparison, direction, and correct units.
- 08
Limitations
Printed beside the number, not buried below it.
Three kinds of evidence, kept separate
The label tells you what a number can support before you read the number itself.
Operating evidence
Shows how the workflow ran: attempts, acceptance, aging, dispositions, returned work, and reconciliation. It is not a patient outcome.
Modeled economics
Calculated from inputs you supply. The formula and assumptions remain visible, and the output is never presented as a Rely result.
Read moreInspect the underlying record under NDA.
For a workflow you are evaluating, we can walk through the eligible population, AI and navigator actions, returned work, disposition, source reconciliation, comparison, and limitations.
- See the definition and denominator behind the measure
- Separate operating telemetry from outcomes
- Review the limitation beside the result
Bring a measure your team already uses.
We will map the population, comparison, period, source, and limitation needed to report it honestly.