Nyze Privacy Policy.

What stays on your iPhone

Notes, audio, attachments, labels, preferences, and capture history are stored locally on your device. The free plan does not send the content of your notes to Nyze servers.

Keyword search and note capture run on the iPhone itself. Nyze should remain useful when you are not on Pro, when you are offline, or when you simply want a fast private notebook.

When the operating system backs up or syncs data, those operations are controlled by Apple and by your iPhone and iCloud settings. Nyze does not read that backup and does not receive a direct copy of your notes because of it.

What Pro features may send

Some Pro features need connected processing: Ask Nyze, semantic search, auto-labels, and some For You insights. In those cases the app sends only the question, note, or relevant excerpt required to deliver that feature.

Nyze backend is designed as temporary compute. Note text passes through the request to create an answer, label, or insight, but it does not become a permanent server-side database of personal content: we do not store your note content, do not cache it, and do not write it to logs.

The AI providers that process that text may retain it for a limited period for abuse and safety monitoring, under each provider's API policy as listed in the subprocessors section. None of them use your note content to train models, and Nyze has no access to those records.

Pro feature processing runs on servers in the us-east-1 region (United States), operated by Amazon Web Services. This international transfer is necessary to deliver AI functionality and occurs based on your consent when using Pro (GDPR Art. 49(1)(a) / LGPD Art. 33).

Ask Nyze answers show the source notes used. That transparency is part of the product promise: AI should not become a black box when it handles personal memory.

Audio transcription and image context are optional Pro features, disabled until you consent in Settings. To transcribe, the app sends a bounded audio fragment; for an image, it sends a normalized copy without personal metadata. OCR is requested only when analysis identifies relevant text in the image. Nyze does not store the audio or image in its backend, and does not use S3 for OCR.

Purchases, subscriptions, and identifiers

Nyze Pro subscriptions and trials are processed by the App Store through StoreKit. Nyze does not receive or store credit card data.

To validate a subscription, we may use a technical device identifier and Apple-signed purchase proofs. This identifier is used to authenticate access to Pro features and apply cost and abuse limits.

We do not use this identifier for behavioral advertising, data sales, or marketing profiles.

Diagnostics and product analytics

We may receive technical crash reports to fix bugs: device model, iOS version, error time, stack trace, and performance information. These reports are processed by the Sentry service as a technical subprocessor.

Crash reports should not contain note content. If unexpected data appears in a technical log, it is treated as sensitive and used only for diagnostics.

We also record usage events to understand whether the app's flows work — onboarding completion, how many notes were created, where a paywall was opened from, whether Ask returned an answer. These events are processed by PostHog and identified by a pseudonym derived from your subscription ID, never by your name or email.

Note content never goes with them. Titles, labels, searches, Ask questions, URLs, phone numbers, and file names are excluded by a technical rule rather than by convention: keys carrying those names are dropped before anything is sent. Counts are reported in ranges, not exact values.

You can turn this off at any time in Settings → Privacy → Share product analytics. iOS analytics are controlled separately, from your device's own system settings.

Feedback you send

If you send feedback from the app, we store the message you wrote and, when you fill the field in, your contact email. We use both to reply and to decide what to improve.

The message is free text and is stored on our server — unlike your notes, which stay on the device. Avoid putting sensitive information there that you would rather keep off the server.

When you delete your account, we remove the link between the message and you: the contact email and the account identifier are stripped from the record. The message itself may remain, anonymized — it no longer points to a person and still helps improve the app.

Retention, deletion, and rights

Because the main content lives on your iPhone, deleting notes inside the app removes those notes from Nyze local storage. If the device has backups, removal from older copies depends on Apple settings.

Temporary technical data used by connected features is kept only as long as practical to deliver functionality, prevent abuse, and diagnose failures.

Transcripts, image descriptions, OCR text, and derived embeddings stay on your iPhone. You can turn processing off to stop future sends; in Settings → Data, you can also delete only those derived results without deleting the original audio or image.

You can delete your account at any time from Settings → Account → Delete Account. That removes the data associated with you on the backend — device record, identifiers, subscription and analytics data — and unlinks any feedback you sent, as described above.

For privacy questions, deletion requests, or data concerns, contact [email protected].

Regional rights and minimum age

Depending on your region, you may have rights under laws such as GDPR (EU), UK GDPR, CCPA/CPRA (California), and Brazil's LGPD. These typically include access, correction, deletion, portability, and the right to object to certain processing.

Because note content lives on your iPhone, most of these rights are exercised directly inside the app. For technical data handled by the backend (device identifier, Pro feature usage records), email [email protected]; we reply within 15 business days.

Nyze is not intended for children under 13. If you are a legal guardian and identify that a child under 13 has used the app, contact our privacy address so we can remove any associated data.

Subprocessors

To deliver Nyze features, we rely on the following technical subprocessors. Each one receives only the minimum data required for its function and is bound by data protection agreements:

Amazon Web Services (AWS) — backend hosting (Lambda + DynamoDB) in the us-east-1 region (United States). Processes Pro feature requests on an ephemeral basis.

Anthropic — Claude models for Ask Nyze, auto-labels, and insights. Receives the question or note excerpt during processing. Contractually prohibited from using data to train models. May retain content for a limited period for abuse monitoring, under its current API policy.

OpenAI — alternate fallback provider used when Anthropic is unavailable, with the same scope of data sent. Contractually prohibited from using data to train models. May retain content for a limited period for abuse monitoring, under its current API policy.

AWS Textract — conditional text extraction from images. Receives only normalized-copy bytes during the request in us-east-1; it does not receive a persistently stored S3 copy.

Voyage AI — embedding generation (numeric vectors) and semantic rerank. Receives note text or query text to produce the corresponding vector. Does not use content to train models.

RevenueCat — subscription management and entitlement validation. Receives a technical user identifier (UUID) and purchase events. Does not receive note content.

Sentry — technical crash and performance reports. Does not receive note content. Region: EU.

PostHog — product analytics with anonymous events. No personally identifiable information is sent. Region: EU. Retention: 1 year.

Apple — payment processing via the App Store and subscription notifications. Does not receive note content.