How it works

One photo becomes a marked, measured skin report.

Skin Savvy AI looks at the skin itself — not a quiz. Here's exactly what happens in the 30 seconds after a shopper taps Scan.

Seven steps

From photo to routine

The first five steps run on the shopper's own phone. The full photo never leaves it.

1
Find the face
478 points place the eyes, brows, nose, lips, jaw and neck on the photo. The face is scaled so every photo is measured the same way.
On the device
2
Keep only skin
Eyes, brows, lips, nostrils, hair, beard and background are removed. Only bare skin is measured.
On the device
3
Correct the lighting
A smooth model of the light falling across the face is fitted and taken out, so a shadow on one side isn't read as darker skin.
On the device
4
Measure each area
Each concern has its own check — colour difference for dark circles, ridge tracing for lines, small dark pits for pores, bright highlights for shine.
On the device
5
Mark what was found
Every finding is outlined on the photo with a smooth shape and a numbered label beside the face. Nothing is marked without a measurement.
On the device
6
AI double-check
Small crops of skin, with eyes and lips blanked out, are checked by AI. Where it agrees, the finding says so.
Masked crops only
7
Build the routine
Findings and skin type are matched to the store's own products — morning, evening, eye and lip — each with a reason.
Masked crops only

Every finding has its own measurement and colour, so shoppers can see exactly what was found and where.

Dark circles
The crescent under each eye is compared with the mid-cheek: how much darker it is, and how different the colour is.
Under-eye creases
Fine ridges under the eyes are traced as lines and measured in millimetres.
Darkening around the mouth
Skin around the mouth and chin is compared with the cheeks; the darker area is outlined and sized.
Uneven tone
Patches on the forehead and outer cheeks that are darker than the skin around them.
Dark spots
Small, separate spots clearly darker than their surroundings are counted and circled.
Redness
Areas noticeably redder than the rest of the face, after lighting is corrected.
Smile lines
The folds from nose to mouth corner are traced as curves, with their length and depth.
Visible pores and texture
Small dark pits on the cheeks and nose are counted per square centimetre.
Shine
Bright highlights on the forehead, nose and chin, as a share of that area.
Lips
Lip tone compared with the skin, and fine vertical lines that suggest dryness.
Neck lines
Horizontal creases across the neck, when the neck is in the photo.

Measured by the same engine on AI-generated sample faces. Tap a number to highlight it.

Loading the sample report…

Shoppers trust a scan more when they can see what happens to their photo. Here's the whole of it.

Measured on the shopper's phone
Face points, skin masking, lighting correction and every measurement run in the browser.
Only masked skin crops leave it
Small crops with eyes and lips blanked out go to the AI double-check, then are discarded.
Photos are never stored
Only the numbers and the routine are saved, so the shopper can revisit their result.

A scan is a careful look at one photo. It's built to help a shopper choose skincare — not to replace a dermatologist.

One free scan in the demo store. About 30 seconds.