SkinCare × Afrodita

AI skin analysis × Afrodita

Only the skincare your skin actually needs.

Upload a cropped photo of the problem area. Our AI identifies visible concerns, applies dermatologist-reviewed rules and selects the smallest suitable Afrodita routine.

Macro fragment of skin under a clinical annotation grid; one analysed area marked
Fig. 01 — scan fragmentmacro · annotated

Privacy

No face required.

Selection

No generic bundles.

Formulation

No duplicate actives.

Protocol

How it works

  1. 01

    Scan

    input · cropped fragment

    A cropped fragment of the problem area — never a full face. Metadata is removed on-device before anything leaves it.

  2. 02

    Consent

    separate · revocable

    Analysis consent is separate from training consent, and both are revocable. No consent, no scan.

  3. 03

    Analysis

    concern · confidence · severity

    Visible concerns are identified and reported with what the system actually knows: concern, confidence, severity.

  4. 04

    Minimal routine

    2–4 products · every choice explained

    Two to four products, no more. Every choice is explained against the rule that fired and the active behind it.

Governance

Rules activate only after a dermatologist approves them.

Every rule in the engine carries an approval record — who signed it and when. Unapproved rules never influence a recommendation.

Honesty

What works, and what is simulated in this demo.

Works today

  • Live catalog of real Afrodita products with prices and stock
  • Rule-based recommendation engine with compatibility checks
  • Scan persistence, routines and progress tracking

Simulated in this demo

  • Photo concern detection — demo returns a plausible fixed reading
  • Ingredient-label OCR — demo parsing is stubbed

Start with a fragment. End with a routine you can defend.

Scan my skin