Data Sharing

Control whether your de-identified articulation samples contribute to model improvement.

How we protect your data

  • All identifying information — client names, clinician names, session notes, and metadata — is stripped before any sample is stored.
  • Only anonymized phoneme-level acoustic features and IPA transcriptions are retained. Audio recordings are never shared.
  • Shared samples are used solely to improve articulation accuracy. They are never sold or shared with third parties.
  • Participation is never required for clinical use. You can turn this off at any time — your access and features are unaffected.

Questions? Contact us or see our Privacy Policy.