Workflow

Three steps. Structured output.

Every Aimazing session moves through the same three stages. The output is always structured, versioned, and attached to a patient's record.

01

Capture

Upload an existing patient photo, drag a batch, or open the in-browser camera for a fresh clinical capture. Images are encrypted in transit and stored in per-clinic scoped storage.

  • Drag & drop or webcam
  • Any device with a browser
  • Per-clinic scoped storage
02

Inference

The image is passed to the analysis engine. The current live engine runs on Lovable AI Gateway (Gemini Vision). The pilot roadmap replaces this with a fine-tuned vision model hosted on managed inference infrastructure.

  • Live: Gemini vision via gateway
  • Planned: fine-tuned classifier
  • Model version stored per scan
03

Structured output

Findings are normalised into a fixed clinical schema: eight concern categories, severity scores, and a plain-language summary. Everything is written to the patient's visit record and searchable.

  • 8 concern categories
  • Severity 0–100 per category
  • Written to patient record
Not a diagnostic device. Aimazing outputs are AI-generated skin assessments to support clinical judgement. Diagnosis and treatment decisions remain with the licensed clinician.