Privacy Policy

Last updated: 2026-10-07

Who we are

NBuilders ("we", "us", "our") operates the Memori mobile application and the website at joinmemori.com. For privacy questions or requests, contact us at [email protected].

Scope

This Privacy Policy describes how we collect, use, and protect information when you use the Memori app and our website. In-app analytics and crash reporting are described below and apply only when you use the mobile app. The website is described in Our website.

Data we collect

When you use Memori, we may collect:

  • Account information: email address and display name from Google or Apple sign-in. We may store linked sign-in identities so you can sign in with more than one provider when they share the same verified email.
  • Memories and media: photos, videos, audio, notes, titles, tags, documents, links, and optional location data you choose to attach to memories.
  • Photo metadata: when a photo is uploaded, its embedded metadata (EXIF), including the GPS location where it was taken, the date and the camera details, is removed from the photo file. Photos uploaded with earlier app versions may still contain it, and deleting them removes it. A location you choose to attach to a memory is stored separately, as described above.
  • Sharing: when you enable a public memory link or invite someone to a folder, we store the share tokens and invite data needed to deliver that content to the people you choose.
  • Push tokens: device push tokens (Firebase Cloud Messaging, or Expo on older app versions) so we can send the notifications you allow (for example activity or sharing alerts). Marketing pushes stay off by default.
  • Usage and device data: product usage events (for example "memory saved" or "screen opened"), linked to an internal account ID, with platform, app version, and operating system version. These events do not contain the content of your memories.
  • Help search: when you type a question in the in-app help chat, we store a short normalized version of the search text (up to 60 characters), excluding text that looks like an email address, phone number, or link, to improve the answers and fill gaps in them. The search runs on your device, without AI, and the text is linked to an internal account ID like other usage events.
  • Crash reports: crash and error logs from Firebase Crashlytics, with device model, OS version, and app version.
  • Purchase information: subscription status and transaction identifiers from the Apple App Store or Google Play (we do not receive your full payment card details).
  • Support communications: messages you send through in-app support or email.
  • On-device only (not uploaded): folder-lock PIN and biometric unlock preferences stay in secure storage on your device and are not sent to our servers.

Ads. The free plan includes 80 memories. After that, nothing is blocked: you can choose to watch a short rewarded video ad (Google AdMob) to make one more save, or upgrade. Ads are optional: they appear only if you tap to watch one, we never show ads automatically, and paid plans have no ads. To show ads, Google may use your device advertising identifier (Advertising ID on Android, IDFA on iOS) and similar device data. On iOS we ask for your permission first (App Tracking Transparency); if you decline, ads are not personalized with your identifier. On Android you can reset or delete your Advertising ID in system settings. We do not sell your personal data, and we do not send your memories or their content to advertisers.

How we use data

We use your information to:

  • Provide, maintain, and improve the Memori service
  • Authenticate your account and secure access
  • Store and display your memories and shared content at your direction
  • Process subscriptions and restore purchases
  • Respond to support requests
  • Detect, prevent, and address abuse or security issues
  • Analyze usage to improve the product (our own product analytics, not advertising)

If you are in the European Economic Area or United Kingdom, we process personal data on the following bases:

  • Contract (Art. 6(1)(b)): processing necessary to provide the service you request.
  • Legitimate interests (Art. 6(1)(f)): security, fraud prevention, crash reporting, and product analytics, balanced against your rights.
  • Consent (Art. 6(1)(a)): where required, for example optional marketing push notifications (off by default), personalized ads on iOS (App Tracking Transparency), and optional Files on device sync (off by default).

Optional: Files on device (off by default)

If you choose Settings → Files on device and turn sync on:

  • Memori downloads copies of the media in the folder you choose to a Memori folder on your device so you can pick those files from Memori in system file pickers (for example government forms).
  • Copies remain on your device. This does not send your files to any additional third party.
  • Sync runs while the app is open (Wi‑Fi only by default). Locked folders are not included.
  • You can turn off sync anytime; we delete the local Memori folder from your device. Your cloud memories are unchanged.
  • Legal basis: your consent (GDPR Art. 6(1)(a)).

Sharing and processors

We share data only with service providers that help us operate Memori:

  • Google Cloud Run (europe-west3, Frankfurt, Germany): API hosting
  • Neon (Postgres database, eu-central-1, Frankfurt, Germany): account and memory records
  • Google BigQuery (EU region): product analytics events
  • Cloudflare R2 (global network): media storage, encrypted at rest by our storage provider
  • Alibaba Cloud DashScope / Qwen (Singapore): AI processing of saves (see Artificial intelligence)
  • Google AdMob: optional rewarded video ads on the free plan
  • Firebase Cloud Messaging (Google), and Expo for older app versions only: delivery of push notifications using your device push token
  • Firebase Crashlytics (Google): crash reporting
  • Google Maps SDK, Google Places and Geocoding: map display and optional address or place lookup (not generative AI). The search text or map location you use is sent to Google
  • Google and Apple: sign-in and in-app purchase processing
  • Email delivery provider: delivery of support messages you send from the app

Who else can see your content. A public memory link, and the media links inside it, can be opened by anyone who has the link, until you stop sharing. People you invite to a folder can see the memories in that folder.

We do not sell personal information. We do not share personal information for cross-context behavioral advertising (CCPA/CPRA).

International transfers

Your data is processed in Germany (EU) (API hosting, database, analytics), in Singapore (AI processing by Alibaba Cloud), and on Cloudflare's global network (media storage and delivery). Some providers (for example Google and Apple) may also process data in other countries, including the United States. For these transfers we rely on appropriate safeguards where they apply, such as adequacy decisions, Standard Contractual Clauses, and the Israeli Privacy Protection (Transfer of Data to Databases Abroad) Regulations.

Retention and deletion

We retain your data while your account is active.

Trash: deleted memories and folders move to trash. Trash is emptied permanently after 24 hours on the free plan and the 3 GB plan, and after 30 days on the 50 GB and 500 GB plans.

Delete individual content: You can delete specific memories or folders in the app. See our Help page for steps.

Delete your account: Settings → Security → Delete account. This deletes your account, memories, media files in Cloudflare R2, folders, sharing data, AI data, push tokens, and support requests from our production systems. Database backups kept by our provider may contain your data until they expire. You can also request deletion without the app by email. See Help · Delete your account for details.

Analytics events: product usage events are kept for product analytics. They are linked to an internal account ID, not to your name or email. To have them deleted, email [email protected].

Your rights

GDPR (EU/UK): You may request access, rectification, erasure, restriction, portability, or object to processing. You may lodge a complaint with your local supervisory authority.

Data export (portability): On every plan you can download a copy of your data in the app: Settings → Export data. The file (JSON) includes your account, folders, memories, and media links.

CCPA/CPRA (California): You have the right to know what personal information we collect, request deletion or correction, and opt out of sale or sharing. We do not sell or share personal information for cross-context behavioral advertising.

Israel Privacy Protection Law: You may request access, correction, or deletion of personal data held about you.

To exercise your rights, email [email protected] or use in-app account deletion. We may verify your identity before fulfilling requests.

Children

Memori is not directed at children under 13 (COPPA). In the EU, we do not knowingly collect data from children under 16 without parental consent where required by law.

Security

We use HTTPS, access controls, and industry-standard practices to protect your data. Media is encrypted at rest by our storage provider. No method of transmission or storage is 100% secure.

Our website

joinmemori.com is hosted on Cloudflare, which processes your IP address and request data to deliver the site and protect it from abuse. Cloudflare may add a small script that protects email addresses on the page from spam bots. The site loads fonts from Google Fonts, so your browser sends your IP address to Google when the page loads. The site does not use cookies, analytics, or advertising trackers. Your theme and language choice are saved only in your own browser (local storage). Memory and folder share pages load the shared content from our API.

Artificial intelligence and automated processing

Memori uses automated processing, including artificial intelligence (AI), to capture, organize, and display memories. This section is important: it describes third-party AI processing. Save classification, including the title, story, folder, and the other language, is done by Qwen (Alibaba Cloud).

You are interacting with AI

On capture and save as described below, there is no chat with a person. Text and media may be processed by automated AI on our servers and by Qwen (Alibaba Cloud) to classify the save. We disclose this:

  • Before sign-in: the consent line on the login screen names Qwen (Alibaba Cloud) and links to this section. Creating an account accepts the Terms, this Privacy Policy, and this AI processing.
  • Published Privacy & Terms: full AI disclosure on this page and in the Terms of Service.
  • On translated content: a "Machine translation" label on memory detail when you view the non-source language.
  • In the app's Help & support screen: a link to our Help page, which links to this section.

Features that use AI

Feature When it runs What is processed
Capture archivist On a new memory saved through capture or share-in (text, media, documents, audio, and/or a link). If classification fails, the memory still saves with the text you wrote, or a general title if you wrote nothing. If your plan's AI credits are used up, the save does not happen Text, files, optional instructions, text from links and documents, the app language, and short personal facts if any. Sent to Qwen
Audio transcription When a save includes audio or a video with sound The audio is sent to Qwen to produce a text transcript, which is then used like text you wrote
Bilingual titles and stories On an AI save, the other language is written by Qwen. A save that skips classification (for example an offline queue) keeps the same text in both languages, with no extra model call. A manual edit is not sent to Qwen On classification: the text and media go to Qwen. Without classification: nothing extra is sent
Folder suggestion Together with the archivist on those same new-memory saves Same capture content as the archivist, to suggest a folder category
Personal context memory After some AI captures, when the model identifies durable personal facts (e.g. a pet name). Also when a title or story you edit by hand includes a name and a look (for example a pet and its breed), with no model call Up to a few short text facts per capture (capped on your account); relevant facts may be re-sent on later archivist calls to improve organization

Not AI: optional map/location search uses Google Places and Geocoding APIs (address lookup), not generative AI.

AI providers
Save classification Qwen from Alibaba Cloud (DashScope / Model Studio, Singapore). Text and media are sent. The title, story, and the other language are stored on the memory
Our role We decide when to call and what to send. Results are stored on the memory
Models Models can change. A material change will be updated here

We do not send a sign-in password. We do not sell content or use it for advertising.

What is sent

To Qwen, on a classified save: text, photos, videos, audio (for transcription), documents, text from a link, the app language, and short facts already stored (for example a pet's name). Qwen also writes the other language.

A save without classification does not send text to another model. A manual edit of a title or story is not sent. If it includes a name and a look (for example a pet and its breed), that short fact is saved on your account and may be sent to Qwen with later classified saves.

AI does not invent facts. The AI writes the title and story only from what you saved and wrote. It is instructed not to add facts you did not provide. Editing a memory never calls AI.

Sensitive information: A memory can include health, money, or family, if that is what was saved. We do not ask for that, but if it is inside the capture it can be sent to Qwen. Do not save content you are not comfortable sharing with this processor.

What Memori does and does not do with AI

We do We do not
Use Qwen to classify a save (title, story, folder, and the other language) and to transcribe audio Use memory content to train Memori's own AI models
Store AI outputs (titles, stories, translations, folder suggestions) in your account Use AI for credit, employment, housing, or similar legal decisions about you
Show labels when you view machine-translated text Guarantee accuracy of AI output
Let you edit, correct, or delete any memory Share content with AI providers other than Qwen for these features (unless this policy is updated)

AI outputs are suggestions and drafts. Review before sharing memories with others or publishing.

Your choices and withdrawal

  • How you accept: AI processing is disclosed on the login screen before sign-in. Creating an account accepts sending the data described above to Qwen for the purposes described, together with this Privacy Policy and the Terms. There is no separate AI notice on every save.
  • Cannot partially disable core AI: Bilingual translation and the capture archivist are built into memory save flows. To avoid third-party AI processing, do not use those features (or delete your account).
  • Stop processing: Delete individual memories, or Settings → Security → Delete account to remove your data from our systems (subject to backup retention). This stops future AI processing on your account.
  • Rights: See Your rights (access, erasure, portability, objection). AI-generated text in your memories is part of your personal data.

Labeling AI-generated and machine-translated content

  • Drafted titles/stories from the archivist are stored as your memory content; you can edit them before sharing.
  • When the other language is shown, the app labels it Machine translation. On a classified save that text comes from Qwen. Without classification, the same text appears in both languages.
  • We do not present AI drafts as human-written journalism or professional advice.

Accuracy, limitations, and your responsibility

AI-generated titles, stories, folder suggestions, transcripts, and translations may be wrong, incomplete, biased, or inappropriate. They are not medical, legal, financial, or other professional advice. You are responsible for reviewing content before sharing or relying on it.

Storage, retention, and deletion

  • AI-generated content is stored in your memory records and metadata (including bilingual content_locales).
  • Personal context facts remain until you delete related memories or your account.
  • Account deletion removes your memories and associated AI data from our production systems. See Retention and deletion.

Legal bases (GDPR)

  • Contract (Art. 6(1)(b)): necessary to provide capture, organization, transcription, translation, and display features you use.
  • Legitimate interests (Art. 6(1)(f)): security, abuse prevention, and improving reliability of AI-assisted features, balanced with your rights.
  • Consent (Art. 6(1)(a)): where your jurisdiction requires consent for this processing.

Changes to AI features

We may change AI models, providers, or behavior. Material changes will be reflected in this policy and communicated in the app when practicable. Continued use after notice may constitute acceptance where permitted by law.

Not legal advice: This disclosure is for transparency. It does not replace advice from a qualified lawyer for your situation or jurisdiction.

Automated decisions

We do not make automated decisions that produce legal or similarly significant effects concerning you. AI-assisted folder suggestions and drafted text are organizational tools only and do not determine your legal rights or status.

Changes

We may update this policy from time to time. The "Last updated" date at the top reflects the latest version. Material changes will be communicated in the app where practicable.

Contact

NBuilders · [email protected]