Scrubs is an iPhone app built at LA Hacks 2026 that automatically detects and redacts protected health information from healthcare photos before they are shared. Built by a team of UCLA students, Scrubs won the ZETIC Company Challenge by solving a real compliance problem that healthcare workers face every day.
What it solves
Healthcare workers share photos constantly, wounds for remote consultation, charts for care coordination, medication lists for dose confirmation. Most of those photos contain patient identifiers that should never leave the device unredacted. Names on wristbands, MRNs on charts, faces in the frame. Redacting manually is slow, easy to skip, and existing tools are locked behind desktop software or expensive enterprise licenses. Scrubs makes it a one-tap action.
What it does
Scrubs uses on-device AI to automatically scan any healthcare photo and detect sensitive information before it is shared. In one tap, it identifies faces, reads text from wristbands, charts and medication labels, flags everything that qualifies as PHI, and masks it automatically. The clinician never has to manually select or redact anything. The photo is clean and shareable in seconds, with no internet connection required and no patient data ever leaving the device.
How it works
Scrubs uses ZETIC Melange to run face detection and PHI classification entirely on-device. On-device OCR reads text from the photo, a PHI classifier identifies sensitive regions including names, MRNs and dates of birth, and face detection flags identifiable people. All sensitive elements are masked automatically. No patient data is uploaded, stored, or sent to any third party. Everything runs on the iPhone the clinician is already holding.
Models used
These are the models used:
Face detection model via ZETIC Melange
PHI classifier via ZETIC Melange
On-device OCR via Apple Vision
Try Melange: https://melange.zetic.ai