Intelligence Layer
- summarize and explain
- surface patterns and changes
- create risk signals as attention
- prepare documentation proposals
- make planning alternatives visible
AI IN BESTCURA
AI should help people see more. It must never decide what people are allowed to do.
BestCura separates intelligence from authority. AI can compress information, surface patterns and generate proposals. Execution remains bound to state, rules, orders and human professional authority.
WHY THE BOUNDARY IS DELIBERATELY HARD
Probability can be valuable. Responsibility requires more.
A model can say something is likely. It cannot by itself know whether an order is still valid, whether a professional is authorized for this exact action in this facility, whether clinical reality has just changed, or which jurisdiction-bound rule is actually applicable now.
Execution Boundary
- AIInferenceProvenance: downstream_execution_authority = NONE
- no independent authorization of clinical actions
- no replacement of professional authority
- no silent rule selection when context is unresolved
- no automatic rewriting of clinical reality
EXAMPLE
AI detects elevated risk. What happens next?
1 · AI Signal
A model produces a signal with provenance.
2 · Attention
Care Flow makes the signal visible.
3 · Professional
A professional interprets it in real-world context.
4 · Authority
An action still requires order, rule and execution right.
5 · Revalidation
Immediately before execution, current state is checked again.
The difference: AI may raise a question. It may not turn its own answer into an authorized action merely because the score is high.
AI may think faster. It may not own more responsibility.
BestCura uses AI as a tool for people. Responsibility remains where context, professional judgment and real care meet.