WristEcho: All-Day Recorder

WristEcho turns the conversations in your day into text you can search and replay. Tap once on Apple Watch, then lower your wrist. Your iPhone transcribes on the device, marks who said each line and writes up your day every evening. No servers, no sign-up, no API keys. ON APPLE WATCH • Tap the big red button in the app, or add WristEcho to your watch face and start with one tap • While the app is on screen, a red dot and timer show it's recording, with battery level; the complication and system mic indicator show it too • Tap the star to mark a moment for the daily report • Hold Stop for one second to finish. Stopped by mistake? Pick up again right away • Audio goes to your iPhone in 10-minute segments; the watch deletes each one only after your iPhone has it • Stops and saves after a long silence (45 minutes by default) or below 10% battery • Tries to resume after a call or Siri interrupts ON IPHONE • No watch? Record on your iPhone, even when locked, or import audio files • Start or stop iPhone recording, or mark a moment, with Siri, Shortcuts or the Action button • On-device transcription in 10 languages, including Mandarin, Cantonese, English, Japanese and Korean, depending on your device • Knows who's talking: read a short text for 20 seconds and your lines are labeled "Me". Family and colleagues can enroll when present and willing. Or name a speaker; with their consent, WristEcho learns that voice and labels it in earlier recordings too • Search every recording, or ask in your own words, with answers citing the original lines • Every evening, a daily report: topics, timeline, to-dos, promises, starred moments and who you talked with. Every Monday, a weekly report • Play from any line at 1×, 1.25×, 1.5× or 2×, straight on into the next segment • Touch and hold a line to fix a word. Share one segment's audio and text AI SUMMARIES • iPhone 15 Pro or later with Apple Intelligence on: reports and answers come from Apple Intelligence, on the device • Other iPhones with 6 GB of memory or more: optionally download an open-source model (about 1 to 2.3 GB) that runs offline • Or extract mode: the most meaningful sentences, no generative model • AI can be wrong. Check important details against the audio PRIVACY • Everything runs on your own devices. No data collected, no account, no ads, no analytics • Transcripts, voiceprints, reports and to-dos go into your own iCloud or computer backup; audio stays out by default. Lock the app with Face ID • Each transcript keeps a voice feature per speaker. An unknown speaker's is deleted with that segment's audio; removing a person deletes theirs • Stranger-only audio is kept 14 days and audio with people you know 90 days by default. Starred segments keep their audio; text always stays • Export all text or delete everything in Settings. Exported diagnostics contain no recordings, text or names TELL PEOPLE YOU'RE RECORDING WristEcho records the sound around you, including other people. Let them know, and follow the recording laws where you are. PRICING One-time purchase. No in-app purchases or subscriptions. REQUIREMENTS • iOS 26 or later; Apple Watch optional, watchOS 11 or later • English, Simplified Chinese and Traditional Chinese • Apple Watch Series 9 test: about 13 hours of screen-off recording on about 74% battery. Results vary Privacy Policy: https://linfanty.github.io/wristecho/privacy.html Terms of Use: https://linfanty.github.io/wristecho/terms.html Contact: wty2003728@gmail.com