Privacy Policy (Speech Notes

Privacy Policy (Speech Notes - AI Note Taker)

Effective date: April 8, 2026
Contact: sami.ios798@gmail.com

1) Overview

Speech Notes - AI Note Taker (“we”, “our”, “the App”) helps you capture voice notes, convert them to text using Apple speech recognition, and optionally generate structured summaries and action points using an external AI gateway. The app is designed so that your voice notes are stored locally on your device, and AI processing (cloud-based summarization/diarization) is triggered only when you choose those features.

2) Information We Collect

Depending on how you use the App, we may process the following categories of information:

A. Microphone audio (when you record)

When you start recording, the App requests microphone access through iOS.
During recording, audio is captured for the purpose of transcription (including optional “live” captions).

B. Speech transcripts

For transcription, the App uses Apple’s Speech framework (SFSpeechRecognizer / SFSpeechURLRecognitionRequest).
Your speech is converted into text either during recording (live captions) and/or after you stop recording (final transcription).
Whether speech recognition runs fully on-device or also uses server processing can vary based on iOS configuration and Apple’s speech recognition behavior.

C. Voice note files and note metadata (stored locally) When you save a note, the App stores the following inside your app sandbox:

Audio file: stored under the application’s Application Support directory (within an Audio/ folder).
Transcript text: stored as part of the note metadata.
Duration and language selection: stored with the note.
AI output (if generated): stored as part of the note (title, summary, action items, keywords).

Concretely, the App persists notes to local storage (a JSON library file and the audio files). Notes remain in local storage until you delete them.

D. Multi-speaker diarization data (AI-triggered) If you use multi-speaker recording, the App can produce speaker-labeled turns by diarizing your transcript after capture stops. This diarization is performed by the AI gateway (see Section 4).

E. Analytics and crash reports The App uses Firebase Analytics and Firebase Crashlytics to:

understand how features are used (e.g., recording flow, whether AI was requested),
and collect crash information.

The analytics implemented in this App includes events such as screen views and success/failure of AI summarization, but the App does not appear to send the raw transcript text to Firebase in the analytics code paths.

F. Advertising information The App uses Google Mobile Ads (AdMob) to serve ads. Ads are also used as part of access gating for AI requests (for example, rewarded ads may be required before AI summarization runs, depending on remote configuration).

F. Notifications If you enable the daily reminder feature, the App schedules local notifications on your device (no external server is required to send the notification content).

3) How We Use Information

We use the information above to:

provide voice recording and transcription,
save and organize your voice notes locally (folders and search),
display transcripts and AI-generated summaries,
run diarization and summarize voice notes when you choose those features,
show reminders (if enabled),
measure app performance and usage through analytics and crash reporting,
provide monetization/advertising through ad services.

4) AI Processing and Data Transfers

When AI is used (summaries/diarization):

If you request AI summarization (or refresh AI insights), the App sends the transcript text (and selected language) to an external AI gateway over HTTPS.
For diarization, the App sends the transcript and a duration hint to the AI gateway so it can return speaker-labeled segments.

AI gateway endpoint (as implemented):

Base URL: https://sami-ai-gateway.sami-apps.workers.dev/ai
Requests are sent via POST with JSON payloads, and an authentication token is included via the request header.

What is sent to the AI gateway:

For summarization: transcript text + language context (via prompt).
For diarization: transcript text + duration hint (via prompt).

Important note on AI outputs:
AI output can be imperfect. The App encourages you to review titles, summaries, and action items before relying on them.

5) Third-Party Services

We use or may rely on the following third parties:

Apple Speech Recognition: for converting your voice to text during live transcription and/or final transcription.
AI Gateway Provider (external service): for summarization and diarization.
Firebase (Google): for Analytics and Crashlytics.
Google Mobile Ads (AdMob): for ad serving and rewarded ad flows.
Firebase Remote Config (via Firebase): to fetch configuration flags such as which ad types are enabled and whether update prompts should display.

Each third-party service processes data according to its own privacy practices. When you use the features that require these services, you may be subject to those third-party policies as well.

6) Data Retention and Deletion

Local retention:

Saved voice notes (audio + transcript + metadata + any AI output) are stored locally and persist until you delete them.
Deleting a note removes its audio file and removes the note metadata from the local library.

Temporary recording files:

During recording, the App writes audio to a temporary file in iOS temporary storage.
When you save, that audio is moved into app storage.
If you discard/cancel, temporary files are removed by the app logic where applicable.

AI processing retention:

The App sends transcript content to the AI gateway only when you trigger AI features.
Any retention of content by the AI gateway is governed by that provider’s policies and operational safeguards.

7) Security

We store data within your app sandbox on your device. No method of transmission or storage is 100% secure. When AI features are triggered, data is transmitted over HTTPS, and we rely on the security measures of the transport layer and the AI gateway/provider.

8) Your Choices

You can control privacy-related aspects by:

Managing microphone and speech recognition permissions in iOS Settings.
Choosing whether to request AI summarization/diarization features.
Deleting saved notes from within the App.
Disabling reminder notifications at the OS level.

9) Children

The App is not intended to collect personal information from children in a manner that would trigger special protections. If you believe a child has provided information through the App, contact us at sami.ios798@gmail.com.

10) Changes to This Policy

We may update this Privacy Policy from time to time. The “Effective date” above indicates when the latest version was published.

Search This Blog

Apps