Embasis Privacy Policy

Embasis is designed around on-device processing. With a supported local model selected, AI inference is performed on your iPhone. Connected features have the separate processing boundaries described below; any optional cloud mode uses its disclosed provider boundary.

Web research

When you use features that require current information, Embasis may make requests to publicly available websites or search services to retrieve relevant web content. Retrieved information is cleaned and provided to the on-device AI model for processing. Your AI inference is not performed by a remote AI service.

Location

If you grant location permission, Embasis may use your location for features such as nearby places, routing, weather, and Smart Notifications. Smart Notifications may use location in the background when enabled to determine when you should leave for a destination. You can disable location access at any time in iOS Settings.

Calendar and reminders

If permitted, Embasis may access calendar and reminder information to display your schedule and create items you request. Access is controlled by iOS permissions and can be revoked in Settings.

Microphone and speech

Meeting features may use the microphone when you explicitly start a meeting session. Saved meeting recordings and transcripts are stored locally. Speech recognition depends on your device, language, and available recognition assets. Where on-device recognition is unavailable, the iPhone speech fallback may send audio to Apple’s speech recognition service, subject to Apple’s privacy practices. On-device language-model inference does not make every speech path local.

Weather and maps

Embasis uses Apple system services such as MapKit and WeatherKit where applicable. Requests to these services are subject to Apple's privacy practices.

Handwriting recognition & Google ML Kit

When you use handwriting recognition features, Embasis utilizes Google ML Kit Digital Ink Recognition. Your pen strokes and recognized text are processed directly on your device and are never transmitted to Google servers for recognition or model training.

In accordance with Google's iOS App Store data disclosure requirements, the ML Kit SDK sends non-identifiable technical telemetry and usage metrics to Google for diagnostics, API performance monitoring, and service reliability. As documented in the Google ML Kit iOS Data Disclosure, this may include device hardware specifications, OS version, app bundle ID and version, API latency, error codes, configured recognition languages, and a per-installation identifier. This technical telemetry is not linked to your personal identity or user account.

Data sharing

Embasis does not sell your personal information. We do not use third-party advertising networks, analytics tracking pixels, or user profiling backends. With a local model selected, language-model inference runs on your device. Web research, maps, weather, downloads, and the speech fallback described above make requests to external services. Any optional cloud mode processes the submitted content under its provider’s disclosed boundary. The selected features, model, and permissions determine which of these paths are used.

Data retention and deletion

Information stored by Embasis remains on your device until you delete it, remove it within the app, or uninstall the app, subject to standard iOS device storage and backup settings you configure.

Your choices

You can revoke permissions such as location, microphone, calendar, and reminders through iOS Settings. You can delete locally stored Embasis content using the controls available in the app.

Children's privacy

Embasis is not intended to knowingly collect personal information from children under the applicable minimum age.

Changes to this policy

This policy may be updated as Embasis changes. The current version will be published on this page.