Methodology
AI estimates, explained.
A Mirra assessment is an AI-generated interpretation of three guided photos and your quiz. It is not a biometric measurement, diagnosis, or objective rating. Lighting, angle, expression, camera, and model output can change the result.
01 — What the model receives
Three photos, interpreted by AI.
Mirra sends the guided photos and relevant quiz context to third-party AI services. The models return structured appearance estimates in eight categories. Those estimates are visual interpretations, not independently validated proportions or lab-grade measurements.
02 — The eight feature reads
Every read, named.
Each scan requests the same eight categories. Available reads may receive a directional label such as strength, maintain, or focus; low-quality or uncertain inputs can reduce the detail returned.
- Skin overallAI estimate of visible texture and tone evenness across the submitted photos
- Under-eyesAI estimate of visible shadow and freshness in the submitted lighting
- BrowsAI estimate of visible grooming, shape, and density
- Hair condition & style fitAI interpretation of visible condition and current styling
- Lips conditionAI estimate of visible hydration and texture — not size or shape
- GroomingAI interpretation of visible grooming details and presentation
- Color harmonysuggested color relationships based on the submitted photos
- Contrast levelAI estimate of the visible hair–skin–eye relationship
03 — How we score
Changeable factors only.
The Glow Score (0–100) is an AI visual estimate derived from the numeric estimates returned for available appearance categories. It can vary even when your appearance has not changed. Treat it as a directional snapshot, not objective truth or a measure of worth.
04 — Your after
How the visualization is made — and checked.
The visualization is generated from your front photo using instructions tied to the plan. Automated AI review checks for obvious identity drift and excluded edit categories. These checks reduce risk but cannot guarantee identity or edit accuracy. The image is an illustrative AI-generated visual estimate, not a prediction; results vary, and every shared image should retain its AI label.
Prompt may request
- Hair styling and suggested color
- Brow grooming and visible density
- Visible skin clarity and tone, conservatively prompted
- Visible under-eye freshness
- Visible lip hydration
- Overall grooming presentation and light
Prompt excludes
- Bone structure — nose, jaw, cheekbones, face shape
- Eye and lip size or shape
- Weight
- Age
- Surgery-like or invasive changes
05 — Evidence status
What this method does not establish.
The Glow Score has not been clinically validated and is not proven to predict real-world outcomes. This page describes the product’s current calculation and safety rules; it does not claim that the score is a scientific health measure or evidence of efficacy.
Mirra provides aesthetic and wellness insights only, not medical advice. Home · Privacy · Terms