Goeckoh is built to supplement clinical speech-language pathology, not replace it. This page is the specific, practical detail a clinician needs to evaluate that claim: exactly what data gets recorded and how, how remote monitoring actually works between sessions, what a correction profile changes, and how to get clinical or district pricing. For the underlying mechanism and evidence, see the Science and Research pages.
The same real-time mechanism a guardian uses, applied to a clinical caseload.
A patient's device generates a short-lived session code each time Goeckoh is in use. Enter that code on your own device, and you get a live, direct connection to that session's acoustic readout — formant positions, voice quality measures, and fluency events — updating in real time while the patient is speaking, wherever they actually are: at home, at school, anywhere the correction is running. You are not limited to observing what happens inside a scheduled appointment.
The connection is a direct relay between the two devices — the server forwards data to whichever monitoring devices are currently connected with the matching code, and keeps no copy of it. It does not accumulate a server-side archive of patient sessions you query later; each live view exists only while that specific session is running. Structured history that persists across sessions is described in the next section.
Structured, countable events — not a subjective summary.
Every session logs discrete, operationally defined events, consistent with the Applied Behavior Analysis measurement approach described in full on the Research page. The point is the same one ABA measurement always makes: whether a skill is improving should be answerable from data, not from how a session felt.
| Field | What it captures |
|---|---|
| Phonological targets | Which vowel or consonant productions were attempted, and how closely each landed to its acoustic target for the active correction profile. |
| Fluency events | Discrete moments — hesitations, self-corrections, successful productions — logged as countable events rather than an overall impression. |
| Voice quality metrics | Formant positions (F1/F2/F3), harmonics-to-noise ratio, and related acoustic measures per utterance. |
| Co-regulation markers | Session-level indicators relevant to emotional/behavioral state during the session, where applicable. |
This structured history is stored in the browser's local storage on the patient's own device — the same architecture described in the On-Device Architecture section of the Science page. It is available for review on that device across sessions, so a trend across weeks is visible, not just a single session's snapshot — but it is not aggregated into a server-side clinical record you query remotely outside of a live monitoring connection.
Different conditions distort the vowel space differently. Profiles aren't a cosmetic setting.
As explained in the "What Formant Correction Actually Changes" section of the Science page, the correction engine shifts a speaker's current formant positions toward a target vowel region. A correction profile determines which target region and how strongly to correct toward it — because the acoustic signature of, say, dysarthria following a stroke is not the same as an ASD-associated speech pattern or apraxia, and correcting toward the wrong target would work against the patient rather than for them.
Calibrated for speech patterns associated with autism spectrum presentations.
Targets the centralized, compressed vowel space typical of dysarthric speech following stroke or neurological injury.
Calibrated for the articulatory-planning signature characteristic of apraxia of speech.
Addresses speech changes following traumatic brain injury, which can present differently from stroke-related dysarthria.
Selecting a profile is the patient's or caregiver's action inside the app — it doesn't require clinician configuration to change — but as the clinician of record, you're the one best positioned to advise which profile actually matches the presentation you're treating, and to review the session data (above) to judge whether it's producing the acoustic change you'd expect.
Designed to extend a therapy plan you already own, not compete with it.
Goeckoh is not a therapy protocol in itself — it doesn't prescribe a treatment plan or make clinical decisions. It's a real-time correction and structured-measurement tool you can point at a specific acoustic goal you've already identified for a patient, and then review the resulting data alongside whatever standard clinical documentation your practice already uses. Many clinicians use it as the thing that's present in the gaps a weekly session can't reach — a car ride, a classroom, a dinner table — rather than as a replacement for direct clinical contact.
On data handling: no voice audio is ever transmitted to or stored on Goeckoh's servers, under any use case described on this page — the architecture is the same on-device design described throughout this site, not a special clinical mode. We're not making a specific regulatory-compliance certification claim here; if your practice or district has a formal compliance review process, we're glad to work through it directly — contact care@goeckoh.com.
Standard pricing is $20/month per person. For clinics or school districts managing multiple patients, contact us directly for volume pricing — care@goeckoh.com.
Try the free demo, or reach out directly to discuss clinical or district pricing for your caseload.