One-line pitch
TrustNotes turns messy, subjective skincare experiences into structured, searchable routines — and answers real product questions using community evidence with citations, not hype.
What it is
A consumer app where people capture structured routines + reviews (skin type, concerns, budget, timeline, products, steps, tips), explore comparable routines, and ask an AI assistant that responds only when it has evidence from TrustNotes notes.
The problem
Skincare advice is everywhere, but it’s hard to trust or compare:
- “Worked for me” is meaningless without context (skin type, concern, routine, timeline, budget).
- Search returns scattered posts, not decision-ready comparisons.
- Generic AI can sound confident without evidence.
The solution
TrustNotes makes experiences comparable and AI answers grounded:
- Structured capture: routines and reviews with the context that makes them reusable.
- Discover + compare: filters across skin type × concern × budget × timeline; ranking weights completeness, votes, and recency.
- Community signals: “Worked / Didn’t work / Haven’t tried” adds lightweight, honest context (not medical proof).
- Evidence-grounded AI: classifies intent → retrieves matching notes → answers with confidence, evidence count, and linked source notes; refuses/clarifies when evidence is weak.
Why it’s different
- Trust is a product feature: citations, confidence UX, and refusal behavior.
- Retrieval before generation: the model doesn’t freestyle; it answers from retrieved routines.
- Routines > freeform notes: routines bundle the context needed for comparison.
What’s live today (MVP)