Privacy Policy

For the iOS app “声桥 (Shengqiao)” · Effective 2026-11-01

In short: we collect nothing. Shengqiao has no accounts, no servers of our own, no analytics and no third‑party SDKs. Speech recognition, translation, speech output and meeting notes run on your iPhone by default.

1. Where data is processed

Sound from the microphone is passed in memory to the system speech recognizer in real time. Nothing is recorded, saved as audio or uploaded.
Speech recognition uses Apple's on‑device recognizer (SpeechAnalyzer) for most languages. A few languages the device can't recognize locally (e.g. Cantonese, Russian, Arabic, Thai, Vietnamese, depending on device and iOS version) use Apple's speech recognition service, in which case audio is processed by Apple under the Apple Privacy Policy; the language list marks these.
Translation uses Apple's Translation framework on device. iOS downloads offline language packs from Apple the first time a language pair is used.
Transcripts and notes are stored only on this iPhone (in the app's own database) and leave the app only when you export or share them. Deleting the app deletes them. Meeting notes are written by Apple Intelligence's on‑device model by default.
Speech output uses iOS speech synthesis (including your Personal Voice if you created one), on device.

2. Presentation mode

When you host a presentation, your iPhone sends the recognized text and translations peer‑to‑peer over Wi‑Fi / Bluetooth (Apple MultipeerConnectivity, encrypted) to nearby devices that joined with your room code or QR code. No server is involved. Both sides can see the other device's name. Sending stops as soon as you leave presentation mode.

3. Optional: an AI model you configure

You may connect your own AI service (e.g. DeepSeek, Qwen, Claude, using your own API key) to refine translations, translate languages the device doesn't support, and write notes. This is optional and off by default. Before anything is sent, the app shows what will be sent, to whom and why, and sends nothing until you tap “同意并启用” (Agree and enable):

Sent: recognized text and on‑device translations (with a few previous sentences as context), your glossary and meeting context, and the full transcript when you generate notes;
Never sent: audio, your identity, location or contacts;
Text goes directly from your iPhone to that provider, not through us, and is handled under the provider's privacy policy. API keys stay in this iPhone's Keychain.

You can withdraw consent or remove the provider at any time in 我的 (Me) → AI 模型 (AI model).

4. Permissions and why

Microphone	Hear you and the other person to recognize and translate speech live. Nothing is recorded or uploaded.
Speech Recognition	Turn speech into text; a few languages use Apple's speech recognition service (see section 1).
Local Network	Send translations to nearby listeners in presentation mode, or receive them from a speaker.

You can turn any permission off in iOS Settings at any time.

5. What we don't do

No collection, storage or upload of personal data, audio or transcripts;
No analytics, advertising, crash‑reporting or other third‑party SDKs;
No sign‑up and no servers of our own.

6. Please be considerate

Ask for consent before recognizing or translating other people's speech in meetings, calls or public places, and follow local laws.

7. Children

The app is not directed at children under 13 and collects no information from anyone.

8. Changes

If this policy changes we will update this page and the effective date above.

9. Contact

Shenzhen Quanjing Digital Intelligence Technology Co., Ltd. · bailutingyu@gmail.com