Privacy Policy

Overview

EdgeNote AI ("we", "our", or "the app") provides voice recording, transcription, and AI-powered summarization applications for mobile devices (iOS, Android) and desktop computers (macOS, Windows). This Privacy Policy explains how we collect, use, and protect your information when you use our applications.

Scope of This Policy

This policy applies to:

• EdgeNote AI Mobile App (iOS and Android)
• EdgeNote AI Desktop App (macOS and Windows)
• EdgeNote AI Website (edgenote.ai)

Information We Do NOT Collect

EdgeNote AI is designed with privacy as a core principle. The following data never leaves your device:

• Audio recordings - Your voice recordings are stored only on your device
• Transcriptions - All speech-to-text processing occurs locally
• AI-generated summaries - Summaries, action items, and insights remain on-device
• Speaker voice profiles - Biometric voice data is processed and stored locally only
• Personal information - We do not collect names, emails, phone numbers, or any identifying information
• Usage analytics - We do not track how you use the app, what you record, or your behavior patterns
• Location data - We do not access or store your location
• Contacts or calendar - We do not access your contacts, calendar, or other personal data

On-Device Processing

EdgeNote AI uses on-device machine learning models for all core functionality:

Speech-to-Text Transcription

Powered by Whisper, an open-source speech recognition model developed by OpenAI. The model runs entirely on your device using optimized inference engines (whisper.cpp). Audio is processed locally and never transmitted externally.

AI Summarization

Powered by open-source large language models (LLMs) that you choose and download. Supported model architectures include Llama, Mistral, Phi, Qwen, and Gemma. All inference runs locally on your device using llama.cpp.

Speaker Detection (Diarization)

Powered by open-source speaker verification models. See the dedicated section below for important information about biometric data handling.

Speaker Profiles and Biometric Data

What We Collect

• Voice embeddings - 512-dimensional numerical vectors derived from audio segments
• Speaker labels - Names you assign to identified speakers
• Confidence scores - How certain the system is about speaker identification

How It Works

Voice embeddings are mathematical representations that capture voice characteristics without storing the actual audio. These embeddings are:

• Generated entirely on your device
• Stored only in your local database
• Never transmitted to external servers
• Used solely to identify speakers in your recordings
• Deletable at any time through the app settings

Your Rights (BIPA, GDPR, CCPA)

Because all biometric data remains on your device:

• You have complete control over this data
• You can delete individual speaker profiles at any time
• You can delete all biometric data by clearing app data or uninstalling
• We cannot access, sell, or share your biometric data because we never receive it

Recording Conversations - Legal Notice

One-Party vs. Two-Party Consent

In the United States:

• One-party consent states: You may record conversations you participate in without notifying others (most U.S. states)
• Two-party/all-party consent states: All participants must consent to being recorded (California, Connecticut, Florida, Illinois, Maryland, Massachusetts, Michigan, Montana, Nevada, New Hampshire, Oregon, Pennsylvania, Vermont, Washington)

International Considerations

Many countries have strict recording laws. In the European Union, GDPR may require explicit consent from all parties. Always research local laws before recording.

Our Recommendation

When in doubt, always inform all participants that you are recording and obtain their consent. EdgeNote AI is a tool; how you use it must comply with applicable laws.

Background Processing

EdgeNote AI may run background processes to provide a seamless experience:

What Background Processes Do

• Continue recording when you switch apps or lock your screen
• Transcribe audio in real-time while recording
• Generate summaries after recording completes
• Process speaker detection to identify participants

What Background Processes Do NOT Do

• Send data to external servers
• Access data from other apps
• Track your location or activity
• Run when you are not actively using the app

Platform-Specific Implementation

Android: Uses a foreground service with a visible notification to indicate active processing. This is required by Android to perform work while the app is in the background.

iOS: Uses system-provided background audio modes for recording continuity.

Desktop: Standard application process; no special background modes required.

AI Models and Third-Party Content

Model Sources

EdgeNote AI uses open-source AI models that you download to your device. Models are sourced from:

• Hugging Face - A public repository of machine learning models

Model Download Process

When you select a model:

• The app connects to Hugging Face's public API to retrieve model information
• Model files are downloaded directly to your device
• No account or authentication is required
• Your search queries are sent to Hugging Face (not associated with your identity)

AI Output Disclaimer

Model Licensing

Each AI model has its own license terms set by its creators. Common licenses include MIT, Apache 2.0, and Llama Community License. By downloading and using models, you agree to comply with their respective license terms.

Crash Reporting (Mobile App Only)

The EdgeNote AI mobile app uses Sentry, a third-party error monitoring service, to help us identify and fix bugs quickly.

What Sentry Collects

• Error stack traces - Technical information about crashes
• Device information - Device model, OS version, app version
• Anonymous session data - When errors occur (not what you were recording)

What Sentry Does NOT Collect

• Your recordings, transcripts, or summaries
• Your IP address or precise location
• Any personally identifiable information (PII)
• Screen recordings or session replays of your activity
• Speaker profiles or biometric data

Data Retention

Crash reports are retained by Sentry for 90 days, after which they are automatically deleted.

Desktop App

The desktop application does not include Sentry or any crash reporting service.

Local Data Storage

All data created by EdgeNote AI is stored locally on your device:

Data Stored Locally

• Audio recordings (WAV files)
• Transcriptions and speaker attributions
• AI-generated summaries, action items, and decisions
• Speaker profiles and voice embeddings
• Collections, tags, and organization data
• App preferences and settings
• Downloaded AI models

Storage Location

Mobile: App-specific storage in your device's Documents folder (iOS) or internal storage (Android). This data is not accessible by other apps.

Desktop: Application data folder specific to your user account.

Your Data Rights

Right to Access

All your data is stored locally on your device. You can access it directly through the app interface. Since we don't have access to your data, there is no need to request it from us.

Right to Deletion

You have complete control over your data:

• Delete individual recordings from within the app
• Delete speaker profiles individually or all at once
• Clear all app data through your device settings
• Uninstalling the app removes all associated data

Right to Portability

You can export your data in standard formats:

• Audio recordings are stored as standard WAV files
• Transcripts and summaries can be shared as text

App Permissions

Required Permissions

• Microphone - Required to record audio. Audio is processed and stored locally only.
• Storage/Files - Required to save recordings and AI models on your device.

Optional Permissions

• Notifications - To alert you about recording status and action items.
• Background Activity (Android) - To continue recording when the app is minimized.

Third-Party Services

EdgeNote AI integrates with the following third-party services:

EdgeNote AI does not integrate with:

• Advertising networks
• Analytics services (beyond Sentry crash reporting)
• Social media platforms
• Marketing or tracking services

Age Requirement

EdgeNote AI is intended for users aged 16 years and older. We do not knowingly collect any information from users under 16. The app processes all data locally and does not require account creation.

If you are a parent or guardian and believe your child under 16 has provided information through our crash reporting system, please contact us at privacy@edgenote.ai and we will delete that information.

International Users

European Users (GDPR)

Because EdgeNote AI processes data locally on your device, you maintain full control as the data controller. The limited crash reporting data sent to Sentry's U.S. servers is anonymized and does not include personal data.

California Users (CCPA)

EdgeNote AI does not sell your personal information. We do not share your data with third parties for marketing purposes. Your recordings and biometric data remain on your device.

Illinois Users (BIPA)

Speaker voice embeddings are stored locally on your device only. We do not collect, capture, purchase, receive through trade, or otherwise obtain biometric identifiers. All biometric processing occurs on-device under your control.

Changes to This Policy

We may update this Privacy Policy from time to time. We will notify you of significant changes by:

• Updating the "Last updated" date at the top of this policy
• Displaying a notice in the app when you update to a new version

We encourage you to review this policy periodically.

Contact Us

If you have any questions about this Privacy Policy or our data practices, please contact us at:

Email: privacy@edgenote.ai