Method
How the AI skin analysis works
This page explains the real pipeline mechanics: what gets measured, which modules run, and where the product’s limits begin.
What the system measures
Acne: visible inflammatory and surface lesion localization.
Pores: visible pore density, prominence, and zone distribution.
Redness: intensity and coverage of tone-uniformity disruptions.
Oiliness: shine and sebum distribution, especially across the T-zone.
Dark circles: under-eye contrast against the user’s own cheek reference.
Complementary modules: face symmetry, hairline/style framing, and skin-age estimates may appear separately from the five core skin metrics.
How the pipeline runs
The uploaded image is decoded on the backend, validated for size, and optionally auto-cropped around the face.
Face-aware masks keep the analysis inside real skin zones rather than background pixels.
The core skin modules run in parallel: spot mapping for acne, texture analysis for pores, tone mapping for redness, shine analysis for oiliness, and under-eye contrast analysis for dark circles.
Complementary appearance modules are handled separately and are not positioned as medical skin diagnostics.
After the visual modules finish, the app assembles derived metrics, a synchronized care plan, and the downloadable PDF report.
How the AI writing works
The language model receives only derived metrics, module summaries, and care-plan context.
The raw uploaded photo is not sent to the external AI report API.
If the external AI call is unavailable, the system falls back to an internal report template.
What the system does not do
It does not provide a medical diagnosis or replace a dermatologist.
It does not rate attractiveness, symmetry, or ‘face value’ on this page.
It does not recommend sponsored products or run an affiliate skincare funnel.
Module deep dives
If you want narrower landing content, use the dedicated pages for each signal.