Privacy Policy

Who we are

Langoust is a language-learning app operated by the individual operator identified in our Legal notice, established in Switzerland, who is the data controller for the personal data described in this Privacy Policy. Switzerland benefits from a European Commission adequacy decision, and we apply the GDPR to EU users alongside the Swiss Federal Act on Data Protection (FADP). For privacy questions or requests (including account or data deletion), contact us at privacy@langoust.com. For bugs or general support, write to support@langoust.com.

Personal data we process

We process the account, profile, and learning information needed to provide the app, including your sign-in account, display name, language choices, learning goals, tutor settings, saved vocabulary, review history, memory notes, tutor conversations, and app usage needed to run your learning sessions.

Because Langoust is currently for adults, we store your explicit declaration that you are 18 or older or under 18, with its policy version, server timestamp, and submission source. You make this declaration in the first step of sign-up, or later on the age confirmation page. If you answer that you are under 18 during sign-up, we do not store that answer and sign-up stops; you can then sign out or delete the account created when you signed in. We use this record to control access to AI learning features. We do not collect a birth date, identity document, or biometric data for this check.

If you enable daily practice reminders, we store your reminder settings, timezone, browser permission status, and browser push subscription. If you allow optional product analytics, we process pseudonymous feature events and account-linked counts of completed onboarding, conversations, practice, explicit vocabulary saves, learning results and AI costs. We may also record the named signup campaign from the link you followed and your membership of an operator-managed pilot group.

We also process technical logs and error reports needed for security, debugging, abuse prevention, and reliability. We avoid intentionally sending sensitive learning content to analytics, and error reports are scrubbed before being sent where the app controls that processing.

Why we use it

We use your data to create and secure your account, personalize your language-learning experience, save vocabulary and review progress, generate tutor replies and explanations, provide speech or audio features, and maintain the reliability and security of the service.

For a Live Tutor voice session, our server opens the session with OpenAI for you and keeps the provider’s call identifier with a hang-up record, so the session can be ended when you end it, when it reaches its time limit, when your age declaration changes or when you delete your account. This record contains no transcript or audio; the conversation audio flows directly between your browser and OpenAI.

Our legal bases are: performance of the service you ask us to provide; our legitimate interests in security, debugging, fraud prevention, and service improvement; consent for optional product analytics; consent and browser permission for daily practice reminders; and legal obligations where record keeping is required.

We do not use automated decisions that produce legal or similarly significant effects for you.

Service providers

We use service providers to run the app, including hosting and database providers, authentication providers, AI and audio providers, payment processing, error monitoring, optional analytics, email delivery, and browser push notification providers. Depending on the feature and deployment configuration, these may include Google services, Neon, OpenAI, Mistral, Stripe (payments), Sentry, PostHog, Resend (email delivery), and browser push services.

These providers process data only as needed to provide their services to the app. Some providers may process data outside your country or the European Economic Area. Where required, we rely on the provider’s applicable transfer mechanism, such as an adequacy decision or standard contractual clauses.

Payments

If you buy a plan or a one-time item, our payment provider Stripe processes your payment details — we never see or store your full card number. Stripe collects your billing address and card country to process the payment and, where applicable, determine VAT. We keep purchase records (product, amount, status, and Stripe reference identifiers) for billing history, legal record-keeping, and fraud prevention. If you delete your account, we retain a pseudonymized usage snapshot (a hash of your email and recent usage totals) for up to 60 days strictly to prevent quota abuse. This is pseudonymous data, not anonymous data, and is deleted automatically.

Reminders and analytics

Daily practice reminders are optional. We only request browser notification permission after you choose to enable reminders. You can turn reminders off anytime in Settings.

Product analytics are optional and separate from account setup and reminders. You can use the app without allowing them. After you consent, our operator dashboard uses activity recorded from the start of your current consent period to measure activation, usage, return visits, practice results and costs. Reports can be filtered by a named signup campaign or pilot group; employers do not receive account access or individual learning records through this feature. Conversation text, saved words, audio and personal notes are not included in these analytics reports.

If you enable “Help improve Langoust”, PostHog receives a limited set of feature-usage events. These events use technical identifiers and, after your consent, may be associated with a pseudonymous Langoust user identifier, so that we can understand product usage across sessions. We do not send your email address, name, conversations, saved vocabulary, audio, prompts or transcripts to PostHog. Attribution information such as UTM source, medium or campaign may be included; advertising click identifiers are not sent. Your browser stores your analytics preference. When analytics are enabled, PostHog may use a pseudonymous identifier linked to your Langoust account. Disabling analytics stops future analytics collection and resets the browser-side analytics identity. A stored preference alone does not enable analytics for another signed-in account. Automatic click and page tracking, typed-text and element capture, and session recordings remain disabled. Campaign attribution uses only a bounded campaign code carried in the signup link; we do not create advertising cookies for it. Attribution is saved only when you first complete onboarding with analytics enabled.

You can withdraw in Settings at any time. Withdrawal stops future optional analytics and removes your account-linked signup attribution and analytics Save records. Existing learning and billing records stay available for their original service purposes but are excluded from product analytics. If you consent again, measurement starts a new period without importing earlier activity. Accepting an updated Privacy Policy does not itself grant analytics consent.

Emails and email tracking

We send account and progress emails (for example, a welcome message and the periodic progress recaps you choose during onboarding) through our email provider, Resend. You control the recap frequency, including turning it off, in Settings.

These emails include click tracking: when you follow a link in one of our emails, the link is routed through our provider so we can see that a link was clicked. Open tracking (a hidden pixel that records when an email is opened) is disabled. You can avoid click tracking by not clicking links in the emails, and you can stop these emails at any time in Settings.

Shared vocabulary content, the leaderboard and your browser

Vocabulary cards have a shared part and a personal part. Shared dictionary-style content may be reused for other learners who save the same word, with language-specific versions where applicable. Your personal example, private note and other learner-specific content stay on your own copy. Content generated using the context of your private conversation is kept learner-specific and is not promoted into shared card content.

The leaderboard is opt-in and off by default. We do not automatically publish your email address or profile name. If you join, the leaderboard shows the display name you choose when joining, your points and your streak, and, only if you choose it in Settings, your photo or your tutor avatar. You can leave at any time in Settings.

The app keeps some data in your browser: settings and preferences in local storage and session storage, kept separately for each account that signs in on that device, and, if you download content for offline use, the downloaded chapters and audio in the browser’s cache storage. Signing out, or deleting your account, removes the account-specific data the app stored in that browser: your settings and offline journals in local and session storage, and the chapters and audio you downloaded for offline use. Offline content is never served to another account. Device-level preferences that are not tied to an account, such as the theme or sound settings, are kept. We do not use advertising cookies.

Shared collections and creator voice

Publishing a collection is optional. We store its selected title, description, cards, languages, link, publication status and times. No creator name is shown on a shared collection. Anyone with a public or unlisted link may view and copy those details. An unlisted link is not a private access control. Other learners can import the selected vocabulary into their own accounts; those text copies, including a folder named from the publication title, can remain after you remove your publication. Voice recordings are not copied into imported vocabulary: a learner who imported the collection can play your recording from the imported card through Langoust, under the same conditions as on the shared page.

If you separately choose to record your own pronunciation or example, your browser prepares a short audio clip and sends it to Langoust only after you preview and confirm it. Our server evens out its volume (loudness normalization and gentle compression) and stores only that processed, compressed copy (Opus audio), not your original recording, in a private Google Cloud Storage bucket in Frankfurt, Germany. With it we store its duration and measured loudness, a hash of the spoken card text, the consent version and time, linked to your account and publication. Clips are limited to 5 seconds for a word or 10 seconds for an example, with a limit of 32 MiB and 1,000 saved clips per creator. We use them to play the voice on the shared collection page; they are not used to clone your voice or train an AI model. No recording plays automatically.

Revoking a collection link immediately stops new access to its page and voice through Langoust, including playback on cards others imported, while retained clips become private and remain available for you to manage or delete. You can make a revoked link active again; the same page, cards and recordings are then available at the same address. Deleting an individual recording stops new playback through Langoust at once, including for people who previously imported the text, and removes the stored audio; the storage provider’s recovery copy is purged within 7 days. Deleting your account removes your source publications, creator attribution on those publications and saved recordings in the same way. We cannot retrieve audio or text copied independently outside Langoust. If personal information appears in text already imported by another learner, contact us so we can assess that specific request; we do not silently delete another learner’s vocabulary.

Retention

We keep your account and learning data while your account is active so the app can preserve your vocabulary, review history, and tutor context. We keep legal acknowledgement records so we can show which Terms of Service and Privacy Policy version you accepted.

We keep the age-declaration record while your account exists so we can apply your current eligibility choice; deleting the account deletes this record with it.

Signup attribution and analytics Save records are kept until you withdraw analytics consent or delete your account; deleting a saved vocabulary item also removes its analytics Save records. Usage reports are recalculated from the current consent period, so withdrawing consent or deleting learning content can change earlier totals. Raw AI usage events are removed after 90 days, and Live Tutor session records and their connection (hang-up) records after 395 days; per-account cost totals remain with the account. A word you started adding but did not finish is deleted 7 days after it expires, and an unfinished checkout is deleted after 30 days; completed purchases are kept as purchase history. Feedback you send us follows the windows stated in its own section above. If you ask us to delete your account, we will delete or anonymize personal data unless we need to keep limited records for legal, security, or abuse-prevention reasons.

Where we keep pseudonymized data, for example a hash of your email address, it is pseudonymous rather than anonymous, and it is deleted on the schedules stated on this page.

Your rights

Depending on where you live, you may have rights to access, correct, delete, export, restrict, or object to the processing of your personal data. Where processing is based on consent, you can withdraw that consent at any time without affecting earlier processing.

You can update your profile in Profile and reminder settings in Settings. For other privacy requests, contact us at privacy@langoust.com. You also have the right to contact your local data protection supervisory authority.

Sensitive information

Please avoid entering sensitive personal information unless it is necessary for your learning context. The app is designed for language learning and is not intended for medical, legal, financial, or other regulated professional advice.

Changes

We may update this Privacy Policy as the product evolves. If the changes are material, we will update the version shown on this page and may ask you to acknowledge the updated policy before continuing to use the app.

Version 2026-09-24 describes adult-only access, the age-declaration record, the Live Tutor connection record, shared collections and optional creator voice, including how links, imported text and recordings are managed or deleted.