Privacy notice for the EUlyra Grow app
This notice tells you which data the EUlyra Grow app for iOS and Android (together with its related web pages, the "app") processes, why, on what legal basis, who receives it, how long we keep it and what rights you have. It applies to the app. Our websites and EUlyra as a platform for companies and clubs have their own notices. If the German and English versions differ, the German version applies.
1. Who is responsible
The controller is
NAVIGRO UG (haftungsbeschränkt), Schmiedegasse 24, 41836 Hückelhoven, Germany, represented by its managing director Tizian Kneschk, registered at the Local Court of Mönchengladbach under HRB 23800, e-mail consumer-support@eulyra.eu ("NAVIGRO", "we").
We have not appointed a data protection officer because the law does not require one. Your contact for data protection is Tizian Kneschk at the address above.
In the app you are NAVIGRO's customer, not your employer. We process your data under our own responsibility, not on behalf of any company or club, and we do not pass it on to employers.
2. The essentials
- For your account we only need your name, your e-mail address and a password or passkey. Everything else (your goal, self-check, plans and learning preferences) you enter yourself, as far as you wish to use the app.
- The app runs on servers in Germany (Hetzner, Falkenstein; backups in Nuremberg). Data on our servers and in backups is encrypted.
- AI features run through Mistral AI in the EU. We never send your name, e-mail address or account identifiers; texts you paste or upload are filtered first, and we show you what was removed.
- The AI does not assess you. You rate yourself; gaps are calculated by fixed code. AI drafts are labelled, and you review and adopt them yourself.
- Statistics collection and push notifications are switched on. You can object to both at any time by switching them off in the app (sections 5 and 6).
- The app contains no advertising, no tracking and no third-party analytics. We do not pass your data to employers, recruiters or advertisers, and we do not sell it.
- You can download your data and delete your account at any time in the app. Deletion becomes final 30 days later.
- The app is for people aged 16 and over.
Beta. Before the official launch, a limited group tests the app. We invite you through Apple's TestFlight or a closed test on Google Play. For the invitation we pass your e-mail address to Apple or Google once; what Apple and Google process while you use the test version (for example crash reports you share through TestFlight) is governed by their own privacy policies. Everything else applies as described; your beta data is kept after launch unless you delete your account.
3. Which data we process and why
The legal basis for everything in this section is Article 6(1)(b) GDPR: we need the data to provide the app to you under the contract for the free version set out in the terms of use. Without the details in 3.1 we cannot set up an account; everything else is optional and only limits what the app can offer you.
3.1 Account and sign-in
You create your account in the app through our sign-in service (operated by us at auth.eulyra.app). We process:
- first and last name, e-mail address (it is your username and must be verified)
- your password, stored only as a hash and never in plain text, or a passkey (public key; the private key stays on your device)
- your language choice (German or English)
- time and IP address of sign-ins, failed attempts, password changes and changes to your account
- your account identifier, by which the app and our server recognise you
Sign-in happens in your system's browser with a procedure in which the app never sees your password (OpenID Connect with PKCE). Afterwards you stay signed in: a refresh token is valid for 30 days without use and for at most 180 days, and is kept only in your device's keychain or keystore. With your sign-in the app creates your personal area on our server (a "personal tenant"): it contains only your data, and only you have access.
3.2 Your confirmations
When creating your account you accept the terms of use, confirm that you have read this notice and confirm that you are at least 16 years old. For each confirmation and each later decision (including statistics collection and push notifications) we store: the type, yes or no, the version of the text, your language and the time. If we change the terms of use or this notice, the app asks you to confirm again before you continue. This record serves as evidence (Articles 5(2) and 7(1) GDPR) and is included in your data export.
3.3 Goal, market profile, self-check and gaps
- Goal: you choose your occupation or next step (occupation code from the German classification of occupations, level, region as a federal state or "Germany") and give the goal a title. One goal is free.
- Market profile: market profiles show what employers ask for in many public job advertisements for an occupation. They contain only figures per occupation and no personal data. If you adopt a market profile, fixed code turns it into your skills with a suggested target level, which you can change.
- Your own skills: you can add skills from the NAVIGRO catalogue or create your own (name, type, target level).
- Self-check: you rate yourself per skill. The app stores your rating with its time; earlier states remain as history. Nobody but you sees them.
- Gaps: target minus current level is calculated by fixed code. No language model, no assessment by us.
3.4 Your own requirement profile and CV (AI features)
These features use language models (section 4). You start them yourself, and each has a monthly quota.
- Job text: you paste a job advertisement or upload it as a PDF. We scan the file for malware and read only its text layer; we do not store the file itself. Before sending, a filter removes e-mail addresses, phone numbers, bank account numbers, dates of birth, names following a salutation, your own first and last name, and addresses. You see the filtered text with the removed parts and confirm that you may share it. The filtered text is kept as the source of your draft. The filter does not recognise other people's names without a salutation, so it is best to leave them out.
- Derived from your employer: you enter your company's name and website and your job title. From public information (website, web search, the company's public job advertisements) we create a company profile and a draft of your requirement profile. The company profile contains only public information; we store it without any link to you in a cache for all users of the same company (fresh for 6 months, deleted after 12 months without use). That you work for this company is stored only in your personal area, and we never approach your company because of it.
- CV → skill suggestions: you paste your CV or upload it as PDF or DOCX, together with your experience and previous job titles. After the filter described above, the language model suggests skills only, no levels and no assessment; you confirm each one individually. We do not store your CV. If the file is a scan without a text layer, we keep it briefly encrypted on our server for text recognition (our own service, no third party) and delete it immediately after reading.
- Approval: every draft is labelled as an AI draft and remains a draft until you have reviewed and adopted it. We store what you change compared with the draft.
3.5 Plan, learning preferences, learning content and ratings
- Plan: you enter a start date, duration, hours per week and optionally a budget. For the draft, the language model receives only your goal (title, occupation, level), your gaps, this framework and your preferences, never your name, e-mail, employer or identifiers. The draft and the resulting plan with its steps, your ticks and your changes stay in your area. "Plan by hand" and "Find learning content" work without a language model.
- Learning preferences: learning platforms, price rule, formats, delivery, language, certificate, hours per week, and optionally your town or region for in-person offers. The town or region goes to the search service as part of the search (section 4), nothing else about you does.
- Ratings: if you rate a completed step, your rating only flows into the NAVIGRO catalogue and into suggestions in aggregated form with others (from three ratings, as average and count), never with your name. Learning content that a plan draft found on the web is added to the catalogue without any link to you and is deleted with your account.
3.6 Notifications in the app and e-mails
- Bell: the app shows you notifications (plan ready, reminder of a due step, account notices). We keep them for 6 months.
- E-mails from the sign-in service: verification of your e-mail address, password reset.
- Account e-mails: when you delete or restore your account or it has been deleted; also when we have to inform you about new versions of our texts or the security of your account.
- Reminder e-mails: only if you switch on "Reminders by e-mail too" under "Me" › "Notifications". The e-mail names the due step of your plan. You can switch it off at any time.
Our e-mails are sent by our provider ALL-INKL in Germany from noreply@eulyra.app (section 7).
4. AI features and search: what leaves our servers
- Mistral AI (Paris, France) provides the language models for requirement profiles from job texts or employer details, skill suggestions from your CV, plan drafts and "Find an alternative". We use the European endpoint; Mistral does not store inputs or outputs (zero data retention) and does not use them for training. What is sent is described in section 3: filtered texts, your goal, your gaps, your framework and your preferences, never your name, e-mail address, account identifiers, your self-check as a whole or your employer. Your CV and job text leave our servers only after the filter and after your confirmation.
- Linkup (Fresnes, France) is our search service for learning offers and for public information about a company. It receives only search queries built from topics (for example a skill and a format), your town or region for in-person offers, and for "From your employer" the company's name, website and your job title, without any identifier of you. A search guard made of fixed code and a small model checks every query first and replaces or discards sensitive terms. Linkup stores queries for at most 30 days on our instruction and does not use them to improve its service.
- Providers' websites are fetched by our servers with the page address only and without any details about you, to check an offer.
- Not used: for content from the app we do not use models from Anthropic (Claude) or other providers outside the EU. This is locked in the code. We use Claude only for NAVIGRO's internal work without user data, such as reading public job advertisements for the market profiles.
For every model call and every search we store a log in your area (purpose, model, token count, duration, status, and the query for searches), without the content of the call. It serves quota accounting, cost control and troubleshooting (Article 6(1)(f) GDPR) and is deleted with your account, at the latest after 24 months.
Labelling and decisions: AI content is labelled as such in the app (Article 50 of the AI Act). The models do not assess you, do not rank people and do not take decisions about you. No automated decision within the meaning of Article 22 GDPR takes place (section 11).
5. Statistics collection, with the right to object
What we do. Once a day, fixed code counts across all app users who have not objected how many people have chosen a learning platform or price rule, which occupations, levels and federal states are chosen as goals, which skills are most often open as gaps, and which NAVIGRO catalogue items have been adopted into plans and completed. We also count searches in "Find learning content" (day, chosen filters and the skill searched for, never your search text). We store figures per group only, with no account, identifier or pseudonym; these figures cannot be traced back to you. Sub-groups (for example an occupation in one federal state) are shown only from 20 people; smaller ones are left out. Your free texts (own platforms, location, goal title) are never counted. We do not count any individual company and do not link anything to employers' data.
Why. We use this to improve the NAVIGRO catalogue and our suggestions (which content proves itself, which gaps are common) and to produce analyses for our marketing, such as an annual skills report. Within NAVIGRO only the people who maintain the catalogue see the figures; reports state the period, the number of people and that app users are not representative.
Legal basis and balancing. Article 6(1)(f) GDPR. Our legitimate interest is to improve the app and the catalogue from actual use and to demonstrate our work with reliable figures. We have weighed this against your interests: only figures per group are stored, never your record; the count runs on data you have entered for the app anyway, with no additional collection and no profile of you; the minimum group of 20 people and the exclusion of free texts and individual companies prevent anyone from recognising you in the figures; and the statistics affect nothing you see or receive in the app. In our assessment your interests therefore do not override ours. You saw this notice when creating your account; statistics collection is switched on from the start.
Your right to object (Article 21 GDPR). You can object at any time: switch statistics collection off when creating your account or later under "Me" › "Consents". From the next daily count you are no longer included. Earlier monthly figures remain; they contain nothing we could attribute to you. An objection stands until you switch collection back on yourself; a new version of this text does not change that.
6. Push notifications, with the right to object
What we do. If you allow notifications on your device, the app registers its token with us. Per device we store: platform (iOS or Android), the token, the app version, the language and when it was last active, for at most ten devices per account. When there is news, we send only a fixed, neutral text through your platform's push service, for example "News about your plan" or "A step in your plan is due soon", together with the type of notification. No plan, skill or name appears on your lock screen; the app loads the content only when you open it. There are three types: plan (drafts ready, plan changed), reminder (due steps) and account. You can switch each type off individually under "Me" › "Notifications". What we have sent is kept for 30 days in an outbox without content.
Recipients. On iOS the Apple Push Notification service delivers the notification (Apple Inc., Cupertino, USA; for the EU Apple Distribution International Ltd., Ireland), on Android Firebase Cloud Messaging (Google Ireland Ltd., Dublin, or Google LLC, USA). Both receive your device token and the neutral text. Apple and Google process these details under their own privacy policies; data may reach the USA in the process. Apple and Google are certified under the EU-US Data Privacy Framework, on which the transfer relies (Article 45 GDPR). We do not place any further push service in between; our use of Firebase is limited to delivery, without analytics.
Legal basis and balancing. Article 6(1)(f) GDPR. Our legitimate interest is that you can use the app meaningfully: a plan depends on you hearing about finished drafts and due steps without opening the app every day. Against this stands your interest in not being disturbed and in not showing content on your lock screen. We therefore keep notifications free of content and infrequent, send them only about your own plan and account and never for advertising, and you keep control through your operating system's permission and the switches in the app. In our assessment your interests therefore do not override ours.
Your right to object (Article 21 GDPR). Switch push notifications off when creating your account or later under "Me" › "Consents": we then accept no further device, delete your registered devices and send nothing more. You can also block notifications in your device settings; the app then receives no token. "Sign out" unregisters your device; "Delete account" removes all devices.
7. Recipients, service providers and security
Within NAVIGRO only the people who operate the app see your data, and even they only when you approve a support case: your personal area is technically locked for our team until you expressly agree in a support case. Every such access is logged and appears in your data export. NAVIGRO's consultants never have access.
We use these service providers as processors under Article 28 GDPR; a data processing agreement is in place with each of them:
- Hetzner Online GmbH, Industriestr. 25, 91710 Gunzenhausen, Germany: servers, file storage and backups in Germany (Falkenstein and Nuremberg)
- Mistral AI SAS, 15 rue des Halles, 75001 Paris, France: language models (section 4)
- Linkup Technologies SAS, 28 avenue des Pépinières, 94260 Fresnes, France: web search (section 4)
- ALL-INKL.COM – Neue Medien Münnich, Hauptstraße 68, 02742 Friedersdorf, Germany: sending our e-mails from noreply@eulyra.app and the mailbox for support requests
We run the sign-in service, malware scanning, text recognition, queues and monitoring ourselves on our own servers. Apple's and Google's push services: section 6. Beyond that, Apple and Google receive nothing from us; what they process when you download and use the app from their stores is governed by their own policies.
Transfers to third countries: we and our processors process your data in the EU. Mistral AI uses providers with locations outside the EU (including the USA) for security and network services; the European Commission's standard contractual clauses apply to these. Push notifications: section 6. No other transfer to countries outside the EU takes place.
Security and logs (Article 6(1)(f) GDPR, our interest in secure, stable operation and in protecting your data):
- For every call to the app's interface, our server records in a log file your device's IP address, the time, the address called, the status code, the amount of data and details of the app or browser. Credentials and tokens are not stored in it. We delete these logs after 14 days.
- To prevent abuse we limit the number of requests per IP address and per account (counters lasting a few minutes).
- The sign-in service logs sign-ins, failed attempts and account changes with IP address for 30 days.
- Uploaded files are scanned for malware on our own servers.
- A change log in your area records what was created, changed or deleted in your account; it is deleted with the account. Our platform log records, without names, when an account was created, scheduled for deletion and deleted (36 months).
- Our application's operating and error logs contain no content and no e-mail addresses; the server overwrites them continuously.
- Connections are encrypted (TLS). Data on our servers is on encrypted volumes, files are additionally encrypted individually, and backups are encrypted.
8. How long we keep data
- Your account and your area exist until you delete the account. It is then blocked immediately and permanently deleted 30 days later: profile, goal, self-check, requirement profiles, plans, learning preferences, consents, devices, logs and your sign-in account. Until the day before deletion you can restore it by signing in, and until then you can still download your data. We send you an e-mail with the deletion date and one when it has happened.
- Inactive accounts: if you have not signed in for 24 months, we notify you by e-mail; if you do not sign in within 8 weeks, we delete your account as described above.
- What remains after deletion: a record of the deletion without name or e-mail address (time, number of files), the identifier of your former area in the platform log (36 months), and the statistics figures, which nobody can attribute to you. Your ratings drop out of the aggregated values.
- Files: we do not store job texts or CVs (section 3.4); a scan is kept only until text recognition, at most one day.
- Company cache (section 3.4): 12 months without use, with no link to you.
- Logs: access logs 14 days, sign-in events 30 days, push outbox 30 days, in-app notifications 6 months, statistics search events 13 months, model and search logs and the change log with the account (at the latest after 24 or 36 months respectively).
- Backups: we back up our database and servers in encrypted form. Backups expire after a fixed period, so deleted data may remain in a backup for up to about five weeks, which we access only in an emergency; after that it is deleted there as well.
There are no statutory retention obligations for the free version; they may arise with a purchase of Premium and will then be set out in the relevant version.
9. What is stored on your device
The app stores on your device only what it needs to work: your sign-in (refresh and identity tokens) in your system's keychain or keystore, never in a cloud backup; a marker of the installation; recently loaded content in memory; an export until you have shared it; and a chosen file until it is sent. On sign-out the app deletes the sign-in, content and export files. For sign-in, your system's browser uses session cookies of our sign-in service, on iOS in an ephemeral session and on Android through Chrome. The app uses no cookies or storage for advertising, analytics or tracking, and it contains no third-party analytics or crash-reporting tools. It asks for one permission only: notifications (section 6). It does not use the camera, location, contacts or photos. In the app switcher the app covers its content.
10. Your rights
You have the right to
- access your stored data (Article 15 GDPR), fastest via "Me" › "Export data" in the app: the file contains everything stored about you in your area, readable as PDF and machine-readable as JSON;
- rectification (Article 16): you change your goal, self-check, plans and learning preferences yourself in the app, and your name and e-mail address through your account in the sign-in service;
- erasure (Article 17): "Me" › "Delete account" (section 8), or write to us;
- restriction of processing (Article 18);
- data portability (Article 20): the export contains your data in a common format;
- object (Article 21): you object to statistics collection and push notifications at any time with a switch in the app (sections 5 and 6). You can object to any other processing we base on Article 6(1)(f) GDPR on grounds relating to your particular situation; we then stop processing the data unless we can demonstrate compelling legitimate grounds.
To exercise these rights, contact consumer-support@eulyra.eu or write to the postal address above. We normally reply within one month. You can also lodge a complaint with a data protection supervisory authority (Article 77 GDPR).
The details for your account are required neither by law nor by contract, but without them we cannot set up an account. Everything else you enter voluntarily.
11. No automated decisions, no disclosure to employers
The app takes no decision that produces legal effects concerning you or similarly significantly affects you (Article 22 GDPR). Language models produce drafts that you review and adopt yourself; they assess neither your suitability nor your performance or personality, and our code discards reasoning that contains such an assessment. Quotas and anti-abuse limits are fixed rules for everyone.
Your data never goes to your employer, to recruiters or to advertisers. A link with an employer's EUlyra access does not exist in this version; if it comes later, it will happen only at your request and under a new version of this notice.
12. Minimum age
The app is for people aged 16 and over. You confirm your age when creating your account. Younger people may not create an account; if we learn of such an account, we delete it.
13. Changes to this notice
If we change this notice, the app shows you the new version with its date before you continue and asks you to confirm that you have read it. You can find the current version at any time under "Me" › "Legal" and on our website.
All versions
- Version of 15 November 2026 (current)