Privacy Policy

Nanari is a private journaling app made by Dong Yoon Lee (“we,” “us”). This policy explains what happens to your information. The short version: your journal is yours, it lives on your device and your own iCloud, and we do not sell it, share it, or use it to advertise to you.

What we collect

Your journal entries, audio, and attached photos. Stored on your device and, if you enable iCloud, in your personal iCloud account through Apple’s infrastructure. We do not have a copy and cannot read them. Photos already in your library that Nanari uses to shape the daily opening question are analyzed on your device only; no image ever leaves your phone.

The full list of what Nanari stores on your device (and mirrors to your iCloud when you enable sync):

• Audio recordings and raw transcripts of your voice entries.
• The cleaned, readable journal text.
• Attached photos.
• Mood ratings and emotion tags.
• On-device embeddings used for semantic search.
• Calendar event titles you attach to an entry.
• Mentions (people, places, pets) extracted from your entries.
• Storyline notes and threads.
• Weather symbol stamped on the entry.
• Place name, only if you turn on location.

Anonymous usage and crash data. To keep the app working and to improve the questions Nanari asks, we may collect app-performance information such as crashes and errors, along with anonymous interaction events, for example which opening-question source was shown and whether it was answered, refreshed, or skipped. This data never includes the content of your entries, your spoken words, or the text of questions Nanari drew from your own life.

Subscription information. Purchases are handled by Apple. We receive confirmation that a subscription is active, but we do not receive your payment details.

How your entries are processed

To write questions and clean your spoken entries into readable text, parts of a journaling session are sent through our secure relay hosted on Cloudflare Workers to two AI services. Both are configured on no-retention tiers: your session data is used only to generate your entry in that moment, is not stored by us, is not retained by the providers for training, and is never sold.

When AI features are on, the following leaves your device:

• To Anthropic (Claude), for language processing: cleaned journal text and small structured signals such as photo tags, calendar titles you attach, mood, weather symbol, and the date-part of the entry. Never raw photos, never exact location, never your photo library.
• To Cloudflare Workers AI (Whisper Large V3 Turbo), for voice transcription: a single voice-recording file per segment when you speak an entry. The audio is transcribed and returned; it is never stored.

Nothing is sent when you journal in Private entry mode, or when AI is off in Settings.

Your off switch. You can disable all AI features at any time from Settings → Nanari Plus (turn off the AI toggle). Journal writing, editing, and browsing continue to work fully offline.

Entries you mark as private are never sent anywhere. They are processed only on your device, with no transcription refinement and no cleanup.

Semantic search runs on your device and does not send your words to us or anyone else.

Weather

Weather data (the icon and temperature stamped on entries) is provided by Apple Weather via WeatherKit. WeatherKit runs through Apple’s own service; no third party sees the request beyond Apple. Attribution: weatherkit.apple.com/legal-attribution.html.

Third-party privacy commitments

Anthropic and Cloudflare are contractually bound to privacy protections equivalent to those in this policy. Both providers operate on zero-retention configurations for our traffic: your data is used only to generate the immediate response and is not stored on their servers, logged, or used to train models. Apple Weather (via WeatherKit) processes queries under Apple’s own privacy protections, which meet or exceed those in this policy, and never receives your exact location.

If any of these providers ever changes its retention or training terms in a way that would weaken these protections, we will migrate off that provider or notify you before continuing.

What we do not do

• We do not sell your data.
• We do not use your journal content for advertising.
• We do not use your entries to train AI models.
• We do not track you across other apps or websites.

Your choices and rights

You can export all of your data as JSON, Markdown, or PDF at any time. You can delete entries, or delete the app, which removes local data (iCloud data follows Apple’s deletion process). Depending on where you live, you may have rights to access, correct, or delete your information under laws such as GDPR or CCPA. To make a request, email us at the address below.

Children

Nanari is not directed to children under 13 (or the equivalent minimum age in your country), and we do not knowingly collect their information.

Changes

If we update this policy, we will post the new version here and update the date above.

Contact

Questions about privacy? Email aiym182@gmail.com.