How it works

You describe what a good photo shows. AI suggests the things to look for, and you edit them. You can add training photos, good and bad, so AI learns what right and wrong look like. Then you set up responses: each one describes a kind of photo, and sets where the learner goes next, their score and the option recorded.

When a learner sends a photo, AI marks each thing to look for as seen, not seen or unclear, then checks your responses from the top:

• A response fits: the learner gets its feedback and goes down its path.
• Nothing fits, like the wrong thing, too dark or too blurry: the learner is asked for a new photo.
• AI isn't sure, or something goes wrong: the photo waits for a reviewer.

Photos are judged on the things to look for, not on looking like your example.