Core principle
The AI in PappAI acts as an operational co-pilot: it proposes, it never decides. The nutritionist retains full professional control at every step.
Where AI intervenes
- Weekly diet plan drafting (AI generation).
- Patient analysis — insights created with
proposedstatus. - Follow-up prioritisation support.
Recommended workflow
Before launching AI
- Patient profile is complete and up to date.
- Care pathway goal is defined.
- Restrictions and preferences are current.
- AI consent is recorded.
During
- Check warnings and outcome status.
- Review
usedModelandfallbackfields. - Avoid unnecessary regenerations.
After
- Verify calorie and macronutrient coherence.
- Confirm the plan respects the patient's real-world constraints.
- Assess practical long-term sustainability.
Operating rules
- Treat every AI insight as an operational hypothesis, not a conclusion.
- Reject generic, redundant, or overly assertive insights.
- Always translate AI suggestions into concrete, verifiable actions.
- Report inconsistent outputs or repeated fallback events as technical issues.