Transparency note
How NutriMama learns
NutriMama improves itself a little each week. Here's exactly what can change automatically, what needs a doctor's sign-off, and what no software can ever touch.
What the brain may change on its own
- Educational content — adding remedies, recipes, FAQs, blog posts, or rewording existing entries to be clearer.
- Small UI polish — copy in buttons and empty states, screen layout, accessibility tweaks.
- AI guidance tone — refining how the chat answers general wellness questions (not medical advice).
These ship automatically only when a 2nd-pass safety audit passes and no clinician on the advisory panel raises concerns.
What needs a doctor's sign-off
- New features that change what data we collect or how reports are generated.
- Medical reasoning updates — changes to how the AI interprets symptoms, lab values, or pregnancy milestones.
- Anything regulated under DPDP — privacy, consent, data retention.
What the brain can never touch
- Payment logic — your subscription, refunds, and billing are never auto-modified.
- Authentication — your account, password, and sessions are off-limits to the brain.
- DPDP consent records — your explicit consent choices stay exactly as you set them.
- Emergency triage — the system that detects severe symptoms and tells you to call 108 is hand-maintained only.
These four are blocked at the code level. Even with full operator approval, the brain cannot edit them.
How clinicians keep us honest
A panel of doctors reviews proposed changes and recent AI answers. Their verdict — approve, concerns, reject, or unsafe — is logged against every change. Any single "unsafe" flag halts the change immediately.
What you can do
Disagree with something the AI said? You can mark it in chat, and a clinician will review it. You can also request a full export of your data, or delete your account, from Settings.