SpeechStart is designed for children. We hold ourselves to the highest standard of privacy protection โ collecting the absolute minimum needed to run the app and nothing more. This policy explains everything clearly.
| Data Type | Purpose | Stored Where | Shared |
|---|---|---|---|
| Child's Name / Nickname | Personalise the experience | ๐ฑ Device Only | โ Never |
| Age Range | Select appropriate exercises | ๐ฑ Device Only | โ Never |
| Exercise Progress | Track improvement over time | ๐ฑ Device Only | โ Never |
| Session History | Build practice streaks & reports | ๐ฑ Device Only | โ Never |
| Voice Input | Prompt practice (not recorded) | ๐ฑ Not Stored | โ Never |
| Face Blend-Shape Data (AR Mouth Coach) | Real-time mouth-shape accuracy & slow-motion replay | ๐ฑ Memory Only | โ Never |
| Purchase Receipts | Unlock premium features | ๐ Apple Only | โ Never |
| Crash Reports | App stability | ๐ Anonymous | Anonymised |
SpeechStart prompts children to practice saying sounds and words aloud as part of the exercises. The app may use the device microphone to let a child tap to record and hear themselves back as a practice aid โ this is purely a local, on-device feature.
We do not record, store, transmit, or analyse audio from your child. No audio ever leaves the device. Microphone access is used only in the moment of practice and is never retained. You can revoke microphone access at any time in iPhone Settings โ Privacy & Security โ Microphone โ SpeechStart.
SpeechStart's optional "AR Mouth Coach" feature uses the device's TrueDepth front camera (via Apple's ARKit face tracking) to show your child where to position their mouth and tongue while practicing a sound. We collect facial blend-shape coefficients (numeric values describing mouth/jaw/lip movement, such as how open the jaw is or how wide the lips are stretched) from ARKit's ARFaceAnchor. We do not collect, save, or have access to a photo, video, or 3D mesh of your child's face โ only these movement coefficients.
Purpose: these coefficients are used solely, in real time, to (1) compare your child's mouth shape against the target shape for the sound being practiced and compute an on-screen accuracy score, and (2) render a colored visual overlay guide. A short rolling buffer (the most recent 3 seconds) is kept only so the app can immediately replay the child's last attempt in slow motion for self-correction.
Storage & retention: all blend-shape data is held only in the device's volatile memory for the duration of the AR Mouth Coach screen. It is automatically discarded โ never written to disk, a database, or any persistent storage โ as soon as the screen closes or a new practice attempt starts. No face data is ever included in session history, progress reports, or PDF exports.
Sharing & transmission: face/blend-shape data is never transmitted off the device, never sent to our servers, and never shared with any third party, advertiser, or analytics provider. ARKit face tracking happens entirely on-device using Apple's frameworks.
Camera access for this feature requires explicit iOS permission and can be revoked at any time in iPhone Settings โ Privacy & Security โ Camera โ SpeechStart, which disables AR Mouth Coach while leaving all other practice features available.
Everything in SpeechStart โ progress data, session history, exercise scores, streaks, and profile information โ is stored exclusively in the app's private sandbox on your child's device. This data cannot be accessed by us, by other apps, or by any third party.
We want to be explicit about what SpeechStart will never do:
SpeechStart includes a dedicated Parent Area (PIN-protected) where parents can:
Any premium features are purchased exclusively through Apple's StoreKit framework. SpeechStart never handles payment information. All purchases require Face ID, Touch ID, or device passcode confirmation โ preventing accidental or unauthorised purchases. Billing is managed entirely by Apple.
Regardless of where you are located, you have the right to access, correct, and delete any data associated with your child. Because SpeechStart stores all data on-device, most of these rights are exercised directly by clearing app data or uninstalling. For any data deletion requests related to anonymised crash data, contact us at aleem.rz@gmail.com and we will respond within 5 business days.
We may update this policy to reflect changes in our app or applicable law. Any material changes will be communicated via an in-app notification shown to the parent account. We will never reduce children's privacy protections without explicit notice. The effective date at the top always reflects the latest version.
We take children's privacy extremely seriously. If you have any concern at all about how SpeechStart handles your child's data, please reach out โ we respond to all privacy enquiries within 1 business day.
aleem.rz@gmail.com