1. The short version
- DRFT never records, stores, or transmits audio.
- iPhone uses the microphone to capture ambient audio (speakerphone calls, in-person meetings). Mac uses system audio capture to listen to meeting apps.
- All speech recognition and wake word detection happen entirely on your device — no audio leaves your phone or computer.
- AI summaries (Pro) send transcript text only to Google Gemini — never raw audio.
- You control when listening starts and stops. The mic is only active while you're running a session.
- The App collects zero analytics . We collect anonymized page views on our website only.
2. Audio capture & microphone access
DRFT captures audio differently on each platform:
- iPhone — DRFT uses the device microphone to capture ambient audio. This is designed for speakerphone calls, in-person meetings, and other situations where audio is audible in the room. You must explicitly start each listening session, and the microphone is only active while a session is running. DRFT does not tap into phone call audio directly — it listens to what's audible around the device.
- Mac — DRFT uses system audio capture to listen to audio output from meeting apps (Zoom, Teams, Google Meet, etc.). You grant the System Audio Recording permission once through macOS System Settings. No microphone access is used on Mac.
On both platforms, audio is processed in real time by on-device speech recognition (Apple's SpeechAnalyzer framework) and is immediately discarded . Audio is never recorded to disk, never uploaded to any server, and never retained after processing.
3. Call recording consent
DRFT does not record calls. It processes ambient audio in real time solely for the purpose of detecting your configured wake words. No audio is saved at any point. However, because DRFT listens to ambient audio that may include conversations, you should be aware of and comply with local recording and consent laws. We recommend informing meeting participants when DRFT is active if required by your jurisdiction.
4. Wake word detection
Wake word detection runs entirely on your device. Audio is transcribed locally into text using Apple's on-device speech recognition, then matched against your configured wake words. No audio or transcript data leaves your device during this process. The speech-to-text stream is processed in real time and discarded — DRFT does not maintain a recording or persistent transcript.
5. AI summaries
When a wake word is detected (Pro tier), a short window of transcript text surrounding the detection may be sent to Google's Gemini API to generate a summary. Key details:
- Only text is sent — never raw audio.
- The text is sent via an encrypted (HTTPS) connection through a Cloudflare Workers proxy that validates requests and enforces rate limits. The proxy does not log or store transcript content.
- We do not retain the transcript text — it passes through our proxy and is forwarded to Google's API.
- Google's data handling is governed by their Gemini API Terms of Service .
If you prefer not to send data to Google, you can use DRFT on the Free tier, which does not include AI summaries.
6. Data stored on your device
DRFT stores the following data locally on your device using SwiftData:
- Missed Moments — Detected wake word, timestamp, transcript snippet, and AI summary (if generated).
- Wake Words — Your configured custom wake words.
- App Settings — Preferences such as detection sensitivity and notification settings.
This data stays on your device unless you opt into iCloud sync (see below).
7. Relay mode (Mac + iPhone)
Pro users can link their Mac and iPhone so detections from Mac system audio are relayed to the iPhone. This uses Apple Push Notification service (APNs) via a secure relay server. Detection alerts are routed through our Cloudflare-hosted relay — only the wake word and a short transcript snippet are transmitted, never audio. Detection payloads (trigger word, timestamp, transcript snippet, and AI summary) are sent encrypted over HTTPS to our relay, then delivered to your iPhone via APNs.
8. iCloud sync
Pro users may enable iCloud sync, which uses Apple's CloudKit to sync detection metadata (wake word, timestamp, transcript snippet, and AI summary) across your devices. No audio data is ever synced. This data is stored in your private iCloud container — we cannot access it. Apple's iCloud privacy practices apply. You can disable sync at any time in the App's settings.
9. Analytics
We collect minimal, anonymized analytics:
- Website — Vercel Analytics collects anonymized page views on drft.ing. No cookies. No personal data.
- App — DRFT does not use third-party analytics SDKs. Usage data such as detection counts, session duration, and achievement progress is stored locally on your device for the gamification system and is not transmitted to any server.
- Advertising — The free tier on iPhone displays ads served by Google AdMob. AdMob may collect device identifiers for ad personalization when you grant App Tracking Transparency permission. No ads are shown on the Mac app or for Pro subscribers.
10. Third-party services
We do not sell, rent, or share your personal data with any third party.
11. Children's privacy
DRFT is not directed at children under the age of 13. We do not knowingly collect personal information from children. If you believe a child has provided us with data, please contact us and we will delete it.
12. Your rights
Depending on your jurisdiction, you may have the right to:
- Access any personal data we hold about you.
- Delete your data — use "Delete All Data" in Settings to wipe all Missed Moments, wake words, and preferences in one tap. You can also delete individual moments from History.
- Opt out of AI summaries by using the Free tier.
- Opt out of iCloud sync by disabling it in Settings.
Because DRFT stores virtually all data on your device, you have direct control over it. Delete the App and your local data is gone.
13. Data retention
Audio is ephemeral — processed in real time and discarded immediately. No audio is ever written to disk or transmitted. Saved Moments (detection metadata and text) persist on your device until you delete them. You can erase all app data at once using "Delete All Data" in Settings. We retain no user data on our servers.
14. Security
All network communication uses HTTPS/TLS encryption. On-device data is protected by iOS/macOS system encryption and your device passcode. Speech recognition runs locally in a sandboxed process — audio never leaves the device.
15. Changes to this policy
We may update this Privacy Policy from time to time. The "Last Updated" date at the top reflects the most recent revision. Material changes will be communicated through the App or this page.
16. Contact
Questions or concerns? Reach out at support@drft.ing .