Safety and automated analysis
Write on paper, and see progress in the work.
The core family flow in APHETRA is: print the materials, the child writes independently on paper, the family uploads the work, the system analyses it automatically, and feedback and follow-up learning tasks are arranged from the performance the work actually shows.
On this page
About AI and human service
Analysis of the work, scoring results and next-step arrangements are produced by automated analysis together with explicit product rules. The product does not promise that a human teacher marks every piece, and “waiting for a human teacher” is not part of everyday teaching. When the work is not yet enough, or different results disagree, APHETRA first withholds a conclusion, asks the family for more material, or switches to a short task to look again.
About information on minors
The family entrance only needs a family credential; a child does not need a separate account. A learning profile can use a code and does not require a real name or school name. Uploaded content is used to complete this family’s learning analysis, feedback and progress records, and for necessary account security and troubleshooting.
What parents can control
Parents can view their family’s learning results and growth record, and can ask for a correction, raise an appeal or request deletion. When we handle a deletion request we confirm the applicant’s identity and keep a record of the handling. Staff who look at a child’s original work need a clear reason and time-limited authorisation.
How to see a child’s progress
A single high score, a successful upload or a single model judgement does not mean ability has steadily improved. We look separately at how this attempt went, whether it still works on a different task, and whether it is remembered after some time. What the work has not yet shown, we tell you plainly.
The public beta helps us find problems in printing, uploading, feedback and follow-up practice, and keep fixing them. A smooth flow does not mean a lasting learning effect has been proven.