sandra.stoerkel

← Back to my path

Privacy policy

1. Controller

The controller for data processing on this website within the meaning of the General Data Protection Regulation (GDPR) is:
Sandra Störkel, C 13-1, 11 Mont Kiara, Jalan Kiara 1, 50480 Kuala Lumpur, Malaysia
Email: sandra.stoerkel@web.de
As I live in Malaysia, Malaysian data protection laws may apply in addition to the GDPR.

2. Summary

  • This website uses no cookies, and no analytics or tracking tools.
  • When you visit, the hosting provider processes technically necessary data (section 3).
  • The connection to the website is encrypted via TLS (HTTPS).
  • Embedded Streamlit applications are only loaded after you actively click to load them (section 5).

3. Hosting and server log files

This website is hosted by Vercel Inc., 440 N Barranca Avenue #4133, Covina, CA 91723, USA.
Vercel delivers the website via a globally distributed server network (content delivery network); the following information applies to it.
When you visit the website, information transmitted by your browser is collected automatically, in particular:

  • IP address
  • date and time of access
  • page requested
  • referrer URL
  • browser type, operating system and device information (user agent)

Purpose: delivering the website and ensuring its stability and security (e.g. defending against attacks).
Legal basis: Art. 6(1)(f) GDPR; my legitimate interest is the secure and error-free provision of the website.
Retention: log data is stored by Vercel according to its policies and deleted once no longer required for these purposes.
Transfer to third countries: Vercel Inc. is certified under the EU-U.S. Data Privacy Framework; the transfer to the USA is based on the European Commission's adequacy decision (Art. 45 GDPR).
More information: vercel.com/legal/privacy-notice

4. Contact by email

If you email me, I process your email address and the content of your message to reply to your enquiry.
Legal basis: Art. 6(1)(b) GDPR where your enquiry relates to a contract or pre-contractual steps (e.g. job application, cooperation); otherwise Art. 6(1)(f) GDPR (legitimate interest in responding).
Retention: I delete the data once your enquiry has been fully dealt with, unless statutory retention obligations apply.
Emails are received via my email provider (WEB.DE).

4a. Contact form

If you use the contact form, I process your name, email address and message to reply to your enquiry. All three fields are required.
How it works: when you submit the form, your input is sent to a server function of this website (Vercel, section 3) and from there sent to me as an email via the email service Brevo. The website itself does not store your message.
Legal basis: as in section 4 (Art. 6(1)(b) or (f) GDPR).
Technical service provider: Sendinblue SAS (brand "Brevo"), 9-17 rue Salneuve, 75017 Paris, France. According to Brevo, data is stored exclusively in the EU (France, Germany, Belgium). A data processing agreement pursuant to Art. 28 GDPR is in place with Brevo (appendix to Brevo's terms of use). Brevo keeps sending logs (including time, delivery status and subject line; the subject contains your name) for one month. Brevo does not store the content of the message as a preview.
Retention: as in section 4.
Spam protection: the form contains a field that is invisible to people. If it is filled in (typical for automated programs), the message is discarded. There is no captcha and no third-party script.
Abuse protection: the server function checks how many messages were sent from your IP address in the last 10 minutes. For this, the IP address is only held in the server function's memory; it is not stored in files or databases and not passed on to Brevo. Legal basis: Art. 6(1)(f) GDPR (legitimate interest in protection against spam and overload).

5. Embedded Streamlit applications (HerEducation, PISA Explorer)

On the HerEducation and PISA project pages as well as on hereducation.sandrastoerkel.com and pisa.sandrastoerkel.com you can load applications running on Streamlit Community Cloud. Streamlit is owned by Snowflake Inc., Suite 3A, 106 East Babcock Street, Bozeman, Montana 59715, USA.

  • The application is only loaded after you click "Load live app". Before that, no data is sent to Streamlit.
  • After the click, your browser connects to Streamlit/Snowflake servers, transmitting in particular your IP address and browser information. Streamlit may use cookies or similar technologies.
  • Legal basis: your consent given by clicking, Art. 6(1)(a) GDPR and Section 25(1) German TDDDG. You can withdraw your consent at any time with effect for the future by leaving or reloading the page; the application will not load again without a click.
  • Transfer to third countries: Snowflake Inc. is certified under the EU-U.S. Data Privacy Framework (Art. 45 GDPR).
  • More information: snowflake.com/en/legal/privacy/privacy-policy/ and streamlit.io/privacy-policy

6. Links to my other applications

This website links to other applications (e.g. Seedwyse, Deutsch-App). They have their own privacy policies, which apply when you use them. Section 6a applies to Reading Adventure.
The reference to my LinkedIn profile is a plain link. No LinkedIn script or widget is loaded; data is only sent to LinkedIn when you click the link.

6a. Reading Adventure (app for children)

Reading Adventure ("Leseabenteuer") is a reading app for children (grades 2–8) at reading-adventure.sandrastoerkel.com and leseabenteuer.sandrastoerkel.com. An AI writes interactive stories in German, English or French; the child decides what happens next. The controller is Sandra Störkel (section 1).

Key points for parents

  • No name, no email address, no password: the app uses an anonymous account.
  • The main character's name stays in the browser on this device and is never sent – not even to the AI.
  • The AI only receives the selections (e.g. world, companion, level) and the story text so far – no account identifier, no name.
  • 30 days after the last visit, the account and the link to its stories are deleted automatically; "Delete everything" does this immediately.
  • No cookies, no analytics or tracking tools, no advertising. Fonts and images are served from my own server.

1. Visiting the app (hosting)
Like this website, the app's pages are served by Vercel Inc.; the technical access data described in section 3 is processed (including IP address, time, browser). Purpose, legal basis, retention and third-country transfer are as described there.

2. Anonymous account (Supabase)
When first needed (e.g. when starting an own story), the app automatically creates an anonymous account with Supabase. It consists of a random identifier; there is no email address, no name and no password. A sign-in key (session token) is stored in the browser so that the child can continue reading next time.

  • Provider: Supabase Inc. (USA). Database and server function run in the Frankfurt am Main (EU) region.
  • When signing in and with every request, Supabase processes technically necessary data such as the IP address and browser information (e.g. to limit sign-ins per IP address and in server logs). For each session, Supabase stores among other things the IP address and browser information; this session data is deleted together with the account (point 6). Supabase deletes the platform's server logs according to its retention periods.
  • In addition, Supabase keeps a technical sign-in log (time, event, IP address, account identifier); its entries are deleted automatically after 30 days. A copy in Supabase's operational logs is deleted after their retention period.
  • Supabase processes this data on my behalf (processor, Art. 28 GDPR; terms: supabase.com/legal/dpa). Supabase Inc. is based in the USA; any transfer there is covered by the safeguards in these terms.

3. What is stored on the server

  • Account: random identifier, time of creation and of the last visit.
  • "My stories": which story parts belong to which of the account's stories and when the story was started and last continued.
  • Story store: the parts written by the AI (text, choices, companion line, puzzle, quiz) together with the selections they were created from (language, world, companion, level, mood, reading level, main character's pronoun) and – if the child pressed a wish button such as "More magic!" before – that wish as a fixed choice. The store contains no name and no account identifier; parts are shown again to other children with the same selections (saves AI calls). A placeholder stands in for the name; only the browser inserts it.
  • Word-help store: the clicked word and its explanation – without an account identifier.
  • Usage counters: for each AI call its type, time, model used, units used, cost and status; while the account exists, linked to its identifier (for the per-account daily limits and the cost brake). Short-term counters for retrievals from the store are deleted after one day.
  • Reports: if a part is reported via "Report", the part, the time and – while the account exists – its identifier are stored; the part is blocked immediately and reviewed by me.
  • Not stored: name, free text from the child, quiz answers, points.

4. What stays in the browser only
The browser's local storage (not a cookie) holds: the main character's name, the last selections, the identifier of the story currently being read, points and reactions, the interface language and the sign-in key. This data does not leave the device (exceptions: the sign-in key is sent with requests to my own server; a chosen wish button is sent to the server as a fixed choice with the next choice, see point 3). "Delete everything" removes it. Safari may remove such data itself after 7 days without a visit.

5. Artificial intelligence (Anthropic)
Story parts, quiz questions and word help are generated by the AI model Claude from Anthropic (USA). The request is made by my server (Supabase function), not by the child's browser. The fixed instructions to the AI are cached briefly (minutes) at Anthropic; they contain no data about the child.

  • Sent to Anthropic only: the selections (language, world, companion, level, mood, reading level, main character's pronoun), the previous AI parts of this story with a placeholder instead of the name, the chosen path, and for word help the clicked word (only words from the story text) and the explanation language.
  • Not sent to Anthropic: name, account identifier, the child's IP address or browser information, free text.
  • Anthropic processes the data as a processor under its commercial terms and, under those terms, does not use API data for training. Anthropic retains requests according to the periods it sets for API customers; transfers to the USA are covered by the safeguards in the data processing agreement.

6. Retention and deletion

  • Automatically: once a day, anonymous accounts whose last visit was more than 30 days ago are deleted – and with them the "My stories" list. Usage counters and reports are then kept without any link to an account (cost and content control). The AI texts themselves stay in the story store without any personal reference. Entries in Supabase's sign-in log (point 2) are also deleted after 30 days.
  • Immediately, yourself: "🗑️ Delete everything" in the app deletes the account on the server and all app data in this browser. If the server cannot confirm the deletion, nothing is deleted and the app asks you to try again. Individual stories can be deleted under "My stories".
  • Usage counters are deleted after 12 months.

7. Error and review logs
The server function only logs technical errors (e.g. AI error code, timeout) and, for newly generated parts, automatic review notes (part identifier, depth, reading level, level, language, type of issue) – no story texts, no account identifiers, no names. Retention according to Supabase's policies.

8. Protection against abuse and costs
The number of new AI parts, quiz questions and word-help requests is limited per account and day; there is also a daily and monthly limit for all visitors together. The usage counters from point 3 are used for this. The app only accepts fixed selection values – free text or instructions to the AI are not possible.

9. Children
The app is aimed at children and should be used with their parents' knowledge. It does not ask for personal data; the name is optional and stays on the device. Unsuitable parts can be blocked immediately via "Report".

10. How the stories are made; French stories
All stories are AI-generated. The openings of the prepared stories were generated in advance and editorially checked (the French ones by an AI assistant, not yet by native speakers). All continuations and other paths are created live and checked automatically (e.g. length, word lists), but are not read by a person before they are shown. Problematic passages can be blocked via "Report".

11. Legal bases

  • Anonymous account, stories, AI requests, deletion: Art. 6(1)(f) GDPR (legitimate interest in providing a free reading offer that a child or their parents choose to use; data minimisation: no name, no email).
  • Usage counters, limits, reports: Art. 6(1)(f) GDPR (protection against abuse and costs, child-appropriate content).
  • Storage in the browser: Section 25(2) no. 2 German TDDDG (strictly necessary for the service requested).

12. Your rights
The rights in section 7 apply. As the app knows neither names nor email addresses, I cannot identify an account without its identifier (Art. 11 GDPR). Quickest way: "My stories" shows the stored stories, "Delete everything" deletes them. For anything else, simply email sandra.stoerkel@web.de.

No automated decision-making, including profiling, takes place.

7. Your rights

You have the right to:

  • access your data (Art. 15 GDPR)
  • rectification (Art. 16 GDPR)
  • erasure (Art. 17 GDPR)
  • restriction of processing (Art. 18 GDPR)
  • data portability (Art. 20 GDPR)
  • object to processing based on Art. 6(1)(f) GDPR on grounds relating to your particular situation (Art. 21 GDPR)
  • withdraw consent with effect for the future (Art. 7(3) GDPR)
  • lodge a complaint with a data protection supervisory authority (Art. 77 GDPR), in particular in the Member State of your habitual residence, place of work or place of the alleged infringement

For any request, simply email sandra.stoerkel@web.de.

8. No obligation to provide data; no automated decision-making

You are not obliged to provide personal data. However, the website cannot be delivered without the technically necessary data described in section 3. No automated decision-making, including profiling, takes place.

9. Last updated

As of 27 October 2026 (go-live). Last updated on 27 October 2026: section 6a "Reading Adventure" added. I will update this policy if the website or the legal situation changes.