Privacy
Updated 10 September 2026
This site asks for no account and no name. Most of what you do stays in your own browser. Below is exactly what the server does process, why, and for how long.
Who is responsible
Evgeny Nikiforov makes and runs this site as a personal initiative and is the controller for your data. Questions or requests: .
On your device
Your progress, results, drafts and transcripts used as answers are saved in this browser and never sent to the server. The Progress page lets you export or delete them; clearing site data in your browser removes everything.
Cookies and browser storage
We use no advertising or tracking cookies and no third-party cookies. The cookies below are needed for the site to work and to stay secure, or have no or negligible privacy impact; that is why no cookie banner asks for consent.
| Cookie | Purpose | Lifetime |
|---|---|---|
inburgering_preferences | Language, level, theme and timer setting | 1 year |
inburgering_position | Your position in the current practice set | 1 year |
inburgering_pass | A random, signed session number the server uses to limit AI feedback and speech recognition per browser. Holds no data about you and is not readable by scripts. | 24 hours |
inburgering_ops | Operator only, after signing in to the operations pages | 12 hours |
| Browser storage | Purpose | Lifetime |
|---|---|---|
inburgering.study.v2 | Your progress, results, drafts and transcripts (browser storage) | Until you clear it |
inburgering.visitor | A random browser number for the anonymous usage statistics | Until you clear it |
inburgering.visited | Counts one visit per browser session (session storage) | Until you close the browser |
Feedback on your answer
When you select Get feedback, the server sends the task, its criteria and your answer text to OpenAI, which generates the feedback. No name, browser number or address goes with it. Speaking feedback uses only the transcript you have checked. Basis: you ask for it yourself by pressing the button; without it nothing is sent. You can always review an answer yourself.
The server keeps the result in memory for up to 30 minutes so a repeated request or a language switch needs no second processing. OpenAI does not use API data to train its models and may keep inputs for up to 30 days for abuse monitoring; we ask OpenAI not to store the response (store: false).
Your voice
Audio stays in your browser while recording. Stopping the recording, or reaching its two-minute limit, sends it to ElevenLabs for Dutch speech recognition. The information icon beside the recording control explains this too, and you can always type instead of speaking. We do not use your voice to identify you.
The server checks the audio in a temporary file and removes it right after validation. Recordings are not stored; transcripts remain in server memory for up to 30 minutes. At ElevenLabs the use of our data for model training is switched off. ElevenLabs may keep a submission for a while under its own policy; zero retention has not been established for this account.
Transfers outside the EU
OpenAI and ElevenLabs are US companies. Transfers rely on the EU-US Data Privacy Framework where the provider is certified and otherwise on the standard contractual clauses in their data processing agreements. There are no other recipients: the site uses no analytics or advertising services and no other external services.
Exercise reports
A report contains only the exercise ID and version, question ID, chosen category, your note and a timestamp. We do not ask for your name or e-mail and do not attach your answer or recording. Do not put personal data in your note; what you write is kept in a database only the operator sees. Use the receipt number to ask for a report to be removed. Resolved reports are deleted no later than a year after they were filed.
Anonymous usage statistics
To see which exercises are hard and how many people practise, the server keeps anonymous events: a random browser number we cannot link to a person, the exercise and question number, the chosen option letter and whether it was correct, the interface language, the level and the time. Your written answers, transcripts and recordings are never part of this. On the server the browser number is kept only in keyed, hashed form; the events are deleted after 90 days.
Security and abuse
To prevent abuse of the paid services, the server counts per network address and per session number how many requests were made in the last minute, hour and day. Those counts live only in the server’s memory, for about a day, and are neither stored nor shared. Per day the server keeps only totals: how many feedback and transcription requests there were, how many failed and how many text tokens were processed. Error messages in the server log contain no answers, recordings or addresses.
Automated processing and AI
Feedback on writing and speaking tasks is generated by an AI model and transcripts by speech recognition. They are aids for practice: not an assessment by a person, not an exam result and not a prediction of passing. They can make mistakes. The site takes no decisions about you and nothing here has legal effect for you; you can always ignore the AI feedback, review yourself, and report mistakes through Report a problem.
Your rights
You have the right to access, correct and erase your data and to object to a processing. Because the server keeps nothing linked to you, you can do most of it yourself: clear the site data in your browser. For a report, use the receipt number. Contact us through the address at the bottom of the page; you can also complain to the Dutch Data Protection Authority (Autoriteit Persoonsgegevens).
This site is meant for adults preparing for the integration exams and is not aimed at children.
Providers and further information
OpenAI privacy policy · ElevenLabs privacy policy · Autoriteit Persoonsgegevens · Terms of use
Changes to this notice appear on this page, with the date at the top.