AestheticSim AI a web app for aesthetic clinics that simulates procedure outcomes with AI
The clinic uploads a patient photo and sees the expected result without changing who the person is.
- 655
- protocol lines
- 7
- Norwood grades
- 7
- skin protocols
- 3
- time milestones
The problem
Patients decide from photos of other people, and generic AI edits change the face itself, which destroys the medical value.
Overview
A multi-tenant web app for aesthetic clinics, each with its own branding. The clinic uploads a patient photo and gets a simulation of the expected result before the patient decides.
What I built
- A 655-line clinical protocol file, not a thin wrapper around the model
- The Norwood scale from NW1 to NW7 with graft-count ranges per grade
- FUE, FUT, DHI and Sapphire techniques with precise milestones at months 3, 6 and 10
- 7 skin protocols with session counts and Fitzpatrick handling
The engineering call
A mandatory rules block forces facial identity to come only from the patient’s photo. The simulation is not allowed to change who the person is; that is the difference between a clinic tool and a photo editor.
