Methodology

How we read a face.

A six-step pipeline built on computer vision, peer-reviewed aesthetic literature and a decade of clinical photographic data.

01

Capture

One front-facing photo, neutral expression, even lighting. We give you a guided capture flow if you need it.

02

Detect

Our model places 478 landmarks across your face — eyes, brows, nasal base, philtrum, jawline, malar peak.

03

Measure

Forty-two ratios are computed: canthal tilt, ogee curve, ramus-to-body ratio, submental angle, lip projection, and more.

04

Score

Scores are calibrated against a reference panel of 50,000 graded faces and 2,000+ academic studies.

05

Plan

An order-of-operations is built — softmaxxing first, hardmaxxing only when ROI clears the bar.

06

Execute

Buy the recommended peptide and skincare protocols, or book a consult with a vetted clinician.

Privacy is not negotiable.

Your photo is uploaded over TLS, processed in an isolated environment, and permanently deleted within 24 hours. We never use your image for training. Ever.

Begin my analysis →