Privacy Statement

This statement explains what personal data the Euro-CNS Deep Learning platform at deeplearning.euro-cns.org processes, why, where it goes, and what rights you have. It describes the platform as it actually runs today. Educational content on the platform (courses, cases, slides) is contributed by Euro-CNS and its partner institutions; this statement covers the platform and your use of it.

Who is responsible

ClearCode .Net Solutions
Melkweg 7, 9718 ER Groningen, Netherlands
info@clearcode.nl

ClearCode operates the platform and is the controller for the data described here. We have not appointed a Data Protection Officer; we are not required to. Write to the address above for anything in this statement.

What we process, and why

Account and profile

Your username, email address, first and last name, and — if you signed up with a password — a salted hash of that password. We never store the password itself. Your profile adds the details you choose to give us: professional status, experience level, institution, department, city, time zone, preferred language, and optionally a phone number. We use this to run your account, to place you in the right courses, and to tell instructors who is in their class.

Sign-in with Google or Microsoft

If you sign in with Google or Microsoft, we never see your password. We receive and store the profile your provider returns, verbatim:

• Google (scopes profile, email): your account identifier, email address and whether it is verified, full name, given and family name, profile picture URL, and hosted-domain if your account belongs to an organisation.
• Microsoft (scopes openid, profile, email, User.Read): your account identifier, display name, given name and surname, mail address and user principal name, and — where your organisation publishes them in its directory — your job title, office location, business and mobile phone numbers, preferred language and age group.

We use the name and email to create and identify your account, and a Microsoft job title, if present, to pre-fill your professional status. The other fields arrive as part of the standard profile response and are kept with the account record. We also store the access token your provider issues, so that a sign-in session can be re-established without asking you to authenticate again.

Learning activity

Course enrolments, progress through lessons and cases, annotations you draw on slides, diagnostic entries and answers you submit, discussion posts and comments, and quiz results. Instructors and platform staff can see the activity of learners enrolled in their courses.

AI Tutor conversations

Described in full in its own section below. This is the most significant processing on the platform and involves sending your messages to an external AI provider.

Uploaded slides and files

Whole slide images and attachments uploaded to the platform are stored in Microsoft Azure Blob Storage, together with the metadata you supply and a record of who uploaded them. Large slide files are sent from your browser directly to Azure and are processed into viewable image tiles by a separate ClearCode-operated service. Slide material must be de-identified before upload — see our Terms of Service. The platform is not designed to hold patient-identifiable data and we ask you not to put any there.

Sign-in security records

We use django-axes to protect accounts against password guessing. It records each sign-in attempt — successful or failed — with the username used, the IP address the request came from, the browser user-agent string, and the time. After five failed attempts for the same username-and-IP combination, that combination is locked out for thirty minutes. These records exist to detect and block attacks on your account.

Technical data

A session cookie keeps you signed in and a CSRF cookie protects forms against cross-site request forgery. Both are strictly necessary; neither is used for tracking or advertising. The platform runs no analytics, advertising or third-party tracking scripts of any kind. Web server logs record IP addresses and request details for security and troubleshooting. If the platform raises an application error, a report is sent to our error-monitoring service (Sentry) — configured so that user identifiers, email addresses and IP addresses are not attached to those reports.

The AI Tutor

What is sent, and to whom

The tutor currently runs on Anthropic's Claude model family, accessed through the Anthropic API. When you send a tutor message, the following is transmitted to Anthropic:

• the message you just typed (up to 8,000 characters);
• the full conversation so far in that case — every previous message from you and from the tutor is re-sent with each turn, because the model has no memory between requests;
• the teaching material for the case: the tutor's instructions, the case history and guided questions you can already see, the hidden teaching notes for that case, and the full text of the reference articles linked to it.

We do not send your name, email address, username or account identifier to the AI provider. The provider receives the conversation, not your identity. The requests are made with a ClearCode API key, so from Anthropic's side the traffic belongs to us, not to you.

The case material is sent using the provider's prompt-caching feature, which holds a copy of that unchanging teaching material on the provider's systems for up to one hour so that repeated requests are cheaper. Your messages are part of the conversation, not the cached block.

The platform also contains an alternative backend for Google Gemini. It is not in use on this installation. If we ever switch to it, we will update this statement before doing so.

What we keep, and for how long

We store your tutor conversations on our own server so you can pick a case back up where you left off, and so we can see where the tutor is helping and where it is failing. Tutor messages are deleted 90 days after they are written. Each message carries its own expiry date and a scheduled job removes expired messages; when a conversation has no messages left, the conversation record goes too. We also count tokens per message to enforce fair-use limits and to track cost.

The first time you open the tutor we ask whether we may keep your chat history. Please read what that question does and does not do:

• It controls our storage. Answering "no" immediately deletes the messages you have written in that conversation from our database, and we will not retain further ones.
• It does not control transmission. Your message has to reach the AI provider for the tutor to answer at all. Declining storage does not make a conversation with the tutor private from the provider. If you do not want your writing to leave the platform, do not use the tutor.

You can also delete a tutor conversation outright at any time from the tutor panel, which removes it and all its messages from our database immediately.

We do not use your tutor conversations to train any AI model, and we do not sell or share them. What the AI provider may do with data sent to its API is governed by its own terms; see Anthropic's commercial terms and privacy centre.

The tutor is a teaching aid and it can be wrong. Nothing it says is a diagnosis or clinical advice.

Who else processes your data

We use the following processors and service providers. We do not sell personal data, and we do not share it for advertising.

The platform server is hosted in Germany. Some of the providers above are established in the United States and personal data may be transferred there, under the transfer safeguards in their respective data processing terms.

Connected learning tools (LTI)

Some lessons embed exercises hosted by an external tool over the LTI 1.3 standard. One such tool is connected today: H5P.com. When you open an embedded exercise, we send that tool a signed token containing your platform user identifier, your full name, given and family name, your email address, and your role in the course (learner). The tool can report a score back to us, and can ask the platform for the membership list of the course context. Your interaction with the exercise takes place on the tool's systems and is governed by its own privacy statement.

Live sessions and recordings

The platform contains a feature for live video teaching sessions and their recordings. It is not enabled on this installation — no video service is configured and no sessions or recordings exist. If we turn it on, we will update this statement first.

Legal basis

• Performance of a contract — running your account, enrolling you in courses, recording your progress, and delivering the education you signed up for.
• Legitimate interests — keeping the platform secure (sign-in protection records, server logs, error monitoring), and understanding how the AI Tutor performs so we can improve it. We have weighed these against your interests and limited the data accordingly.
• Consent — retention of your AI Tutor chat history, and marketing email if you opt in. You may withdraw either at any time without affecting your use of the platform.
• Legal obligation — where we must keep records to comply with the law.

How long we keep things

• AI Tutor messages — 90 days, then automatically deleted.
• Account, profile, enrolments and learning activity — for as long as your account exists. Ask us to delete your account and we will remove them.
• Sign-in security records and server logs — kept as a rolling security record and cleared periodically.
• Uploaded slides and case material — retained as part of the teaching collection, which normally outlives any single account.

Your rights

Under the GDPR you have the right to access your data, to have it corrected, to have it erased, to restrict or object to how we process it, and to receive it in a portable format. To exercise any of these, email info@clearcode.nl. We will respond within one month.

If you are not satisfied with how we handle your request, you can lodge a complaint with the Dutch supervisory authority, the Autoriteit Persoonsgegevens, or with the authority in your own country of residence.

Security

Traffic to the platform is encrypted with TLS. Passwords are stored only as salted hashes. Access to slide storage is granted through short-lived signed URLs rather than shared credentials. Sign-in is rate-limited and locked out after repeated failures. Administrative access is restricted to a small number of named accounts.

Children

The platform is intended for medical professionals, students and educators. It is not directed at children and we do not knowingly collect data from anyone under 16.

Changes

We will update this statement when the platform changes in a way that affects it — in particular if we change AI provider, enable live sessions, or connect a new external tool. The date at the top shows the current version.

Contact

This statement is governed by Dutch law.