Language samples & pronunciation (Pro)

Language samples

Record a natural speech sample — from the student page or mid-session — and get an AI transcript with instant metrics: utterances, total words, NDW, TTR, MLU, and words per minute, plus AI-drafted clinical observations you edit and own. Consent-gated: no recording ever happens without the student’s recording-consent flag.

From any sample, Suggest goals drafts 2–3 measurable goals from the data — pre-filled into the goal editor for you to accept, tweak, or discard.

Sample metrics also flow into evaluation reports and progress reports automatically.

Pronunciation assessment

Recorded trials and home-practice clips can be scored for pronunciation accuracy down to the individual sound — overall accuracy plus a per-phoneme breakdown. It’s framed as trend data, not diagnosis: your perceptual judgment stays the judgment; this puts an objective line next to it over time.

Recordings are captured as 16 kHz WAV specifically so scoring works on every platform.