Powerful AI, little data
We use leading international AI models to understand the update. The audio is sent for processing without additional information about the child or family.
Privacy explained
Your voice is
not a data archive.
Audio is processed, never stored on memdiz servers. The transcript is deleted after subjects, tasks and exams have been extracted.
Nothing to replay
Audio and transcript are not kept as an archive. Once the school day has been structured, they are deleted.
GDPR-compliant
International processing is handled with the safeguards required by European data protection law.
Parents
See the plan, never the recording, transcript or a hidden detail layer.
External AI · firm boundaries
Used to understand.
Never to train.
For speech recognition and sorting, memdiz uses powerful international AI models. We exclusively use access routes with Zero Data Retention and a binding No-Training policy: AI providers do not keep the audio, inputs or results, and do not use them to train their models.
We send no email address, name or additional child profile to these models. They receive only what is needed for that single processing step.