Privacy Policy

Yantra: Second Brain · Last updated October 17, 2026

Yantra is a personal “second brain” app: voice memos, text notes, a daily journal, and saved links — captured, transcribed, embedded, and searched entirely on your device. This policy covers the iPhone, Apple Watch, Android, and Wear OS apps and explains, in plain terms, what happens to your data. The short version: your notes never leave your device except when you yourself ask an AI feature to run, and even then they go directly from your device to the AI provider you chose, using your own API key.

No accounts, no servers, no tracking

Yantra has no user accounts and no sign-in. The developer operates no server that receives, stores, or processes your notes. There is no analytics SDK, no advertising, no third-party tracking, and no data collected about you. Your content is not sold, shared, or used to train any model by the developer.

Where your data lives

All of your notes, journal entries, transcripts, voice recordings, images, saved links, and chat history are stored locally on your device. On iPhone they use Apple’s on-device database and are included in your device’s normal encrypted backups if you have those enabled. On Android they live in the app’s private storage and are not included in Google cloud backups; Android may copy them when you transfer to a new phone, but never your API keys. Uninstalling the app removes this data.

Your API keys

Yantra is “bring your own key.” No key is required: with none, AI features run on-device (Apple Intelligence on iPhone, Gemini Nano on supported Android phones). You can optionally add your own key for Anthropic, OpenAI, or Google Gemini (for chat and summaries) and for Deepgram or Sarvam (for speech). These keys are stored only in your device’s secure storage (the iOS Keychain, or on Android encrypted with a key held in the Android Keystore). They are never transmitted to the developer and are used solely to authenticate your own requests to those providers.

On-device processing

• Microphone & speech.Voice capture is transcribed on-device by default, using Apple’s Speech framework on iPhone and the phone’s on-device speech recognizer on Android. Audio is processed locally and not sent to the developer.
• Embeddings & search. Notes are embedded and searched on-device (on Android with a small multilingual model bundled in the app). Your search queries stay on your device.
• Apple Intelligence.When “On-device” is the chosen model (the default with no key), chat, summaries, tags, and the digest are generated on your device by Apple’s on-device model and nothing leaves your device.
• Gemini Nano.On supported Android phones, the “On-device” choice uses Google’s Gemini Nano, which runs on your phone through Android’s AICore; nothing leaves your device.
• Apple Watch.Voice and text notes captured on the Watch are sent directly to your paired iPhone (Apple’s WatchConnectivity), then transcribed and stored there.
• Wear OS.Voice and text notes captured on a Wear OS watch are sent directly to your paired Android phone through Google Play services’ Wearable Data Layer, then transcribed and stored there.
• Siri & Spotlight.Note titles and excerpts are indexed in your device’s on-device Spotlight index so you can find them from Search and Siri (iPhone only).
• Digest. The daily or weekly digest is assembled on your device. Its written summary is produced by the model you chose (see below).
• Calendar. If you grant calendar access, Yantra reads your selected calendars to show your agenda and writes events you create. Calendar data is processed on-device and is not sent to the developer.

When data is sent to AI providers

If you choose a cloud model in Settings, then when you use an AI feature (chat with your notes, link and image summaries, meeting notes, the digest summary) the relevant content — your question and the matching note excerpts — is sent directly from your device to that one provider, authenticated with your own key. Only the provider you selected receives it:

• Anthropic (Claude). Anthropic Privacy Policy.
• OpenAI. OpenAI Privacy Policy.
• Google Gemini. Google Privacy Policy.
• Apple Private Cloud Compute(optional, off by default, no key) — larger Apple Intelligence requests run on Apple’s Private Cloud Compute servers, which Apple designs so that your data is not stored or accessible to anyone, including Apple. About Private Cloud Compute.
• Deepgram (optional) — higher-accuracy and speaker-labeled transcription. The audio you choose is sent to Deepgram. Deepgram Privacy Policy.
• Sarvam (optional) — Hindi/Indic speech, translation, and meeting transcription. The relevant text or audio is sent to Sarvam. Sarvam Privacy Policy.
• Saved links. When you save a URL, your device fetches that page and summarizes its text with the model you chose (on-device, or the provider above).

Each provider processes that data under its own terms and privacy policy. Because requests use your own account and key, that provider’s data-handling and retention settings apply to you directly.

Backups you control

Yantra can export all of your notes to a single backup file, or as Markdown files (for example, for Obsidian), saved wherever you choose (for example, Files, iCloud Drive, or Google Drive). Backups made on iPhone and Android use the same format, so you can move your notes between them. These exports are created only when you initiate them and are never sent to the developer.

Children

Yantra is not directed to children and does not knowingly collect information from children.

Changes

If this policy changes, the “last updated” date above will be revised.

Contact

Questions about privacy? Email help@singhcodes.com.