Privacy statement
Last updated: October 2026
This is a translation of the Dutch text. In case of any difference, the Dutch version prevails. Read the Dutch version
1. Who we are
TalentProof (“we”) is the controller of the personal data processed through this platform. TalentProof is a trade name of Solutit, Fruitweidestraat 91, 8043 XD Zwolle, registered with the Netherlands Chamber of Commerce (KvK) under number 50836552. Questions? See section 9.
2. What data we process
- Account data: name, email address, role, (encrypted) password or SSO identity, and the language in which you receive emails and notifications (Dutch or English).
- Profile and performance data: the cases you worked on, scores, companies' ratings of your competencies, and swipe preferences. Your solution is assessed by an AI — see section 4.
- If you fill it in: what you currently do (for example student, returning to work or freelancer) and what you are open to (internship, side job, job or assignment). This is voluntary. A company only sees it once the results of its case are final, just like your name. It is not on your public profile, and the AI never receives it.
- If you fill it in when you submit: how you used AI yourself for a solution (which AI, and for what). This is voluntary and does not count. The company behind the case only reads it once the results are final. It is not on your public profile, and the AI never receives it.
- If you indicate it: what you are open to at the company behind a case (internship, side job, job or assignment). This is voluntary and does not count. The company only sees it once the results are final, just like your name; the AI never receives it, and it is not on your public profile.
- The recruitment record between a company and you: when you took part in a case from that company and what you were both open to at the time, the results and from when the company saw your name, when it viewed your solution with your name, contact requests, and whether an internship, side job, job or assignment resulted. This establishes whether an appointment came about through TalentProof (terms, article 8). You can always see the record; the company can see it as soon as it is allowed to see your name.
- If you ask the company a question in a case's Q&A round: your question. The company sees it without your name and can rephrase it neutrally. If it answers it, all participants in that case see the question and the answer, also without your name. So do not put a name or contact details in it.
- For payouts of prize money: address, date of birth and IBAN. No citizen service number (BSN) — we do not ask for it and do not store it.
- Payment data is processed by our payment service provider; we do not store card details.
TalentProof is for people aged 18 and over. You confirm this when you create an account; we do not knowingly process data of minors. If we find out that someone is younger, we close the account.
3. Purposes and legal bases
We use your data to provide the platform (performance of the contract), to assess and match cases, to make payouts and to comply with legal obligations. If the Dutch Tax Administration (Belastingdienst) asks us to report amounts we have paid to you, we provide name, address, date of birth, amount and date — and expressly no citizen service number (BSN). For some processing, we ask for your consent; you can withdraw it at any time.
4. AI assessment of your solution
Your solution is assessed by an AI model from an external provider (Anthropic). This happens when the case closes: the AI then assesses all solutions at the same time, in the last version you submitted. Until then, you can edit your solution. With it, we send the case, the case brief, the company's answers from the Q&A round and your answer — and expressly not your name, email address or any other identifying information. The assessor therefore gets to see your work, not who you are. So do not put any information about your origin, religion, health or other personal background in your solution yourself either: we do not need it and it does not count. Each solution is assessed twice independently, so that we can see whether the judgement is stable.
Based on those scores, our own system gives advice on each solution: not an exact place, but a range, for example “place 2–4”. For this, no data leaves TalentProof. That advice is not the decision: the company itself chooses which solutions it reads — as many as the case has winners, one to five — from the solutions the AI ranks highest (three to eight) or from outside them, and nothing is ticked in advance. The company then also decides the ranking and the winners. We keep both the advice and the company's choice. No decision is therefore taken that is based solely on automated processing. Until the results are final, the company does not see your name: it assesses your work under a number.
The AI also assesses your competencies, using the same fixed description per level (rubric) that companies use. From this, you receive written feedback after the case closes: what the AI saw in your solution, and what a next step would be. If the company does not assess your competencies, you will also see the AI's estimate per competency after the results — only you, and it does not go on your profile. Companies never see the AI's scores. However, when the company assesses each competency, it is shown the parts of your solution that the AI considered relevant to it, without a score or judgement: the company decides for itself whether they are strong or weak. We also use the AI's scores to check whether the AI and companies see the same thing, and as a signal for the participation guarantee we give the company; that guarantee counts the solutions that clear the bar (they make it plausible that they meet at least half of the success criteria). Where necessary, a person at TalentProof decides on this, and it has no consequences for you. If a company reports that your solution on its shortlist does not clear the bar, a person at TalentProof decides as well, not the AI; if it does not clear the bar, it drops off the shortlist, and you will see why after the results. Anything you do not see in the app, you can always request from us. What goes on your profile is the rating of a company that read your solution: before the results are final, it uses that rubric to give a level, or a range of two levels, for each competency, without seeing your name.
While you are still working on your solution, the AI responds to your work once. After you have locked the first version of your solution, that version is sent with the case (without your name) to Anthropic. The AI says what works, what is missing or does not quite fit, and what questions a decision-maker would ask, without a score or judgement. In the step “The AI's take”, you write what you do with it. We store the response with your solution, and the company reads it alongside what you did with it. If the AI cannot respond at that moment, that step may be left empty.
Just before you submit, you can have the AI check whether your solution covers everything the company asks for: what you are asked to submit and each success criterion. You can do this twice. For this, your solution (without your name and without the summary) is sent with the case to Anthropic. For each point, the AI says whether it is covered, not how well. That result is for you only: the company does not see it, and it does not count in the assessment. We store the latest result with your solution.
At the end, you write the core of your advice in one or two sentences; this counts in the assessment. The AI then drafts a summary of your solution once, in two short parts: what you propose and why it works. For this, your solution (without your name) is sent to Anthropic once more. You edit the summary and confirm that it is correct, or you leave it empty. The company sees it next to your core, with a note that the AI drafted it and you approved it. The summary does not count in the assessment, but the company chooses its shortlist based on your core and your summary.
If you disagree with your assessment, you can ask on your own submission for a person to look at it. Someone from our team then reads your solution themselves and you receive the outcome. That is different from running the model again: then the same assessor would look at the same work once more.
The AI provider processes your solution outside the European Economic Area, under a data processing agreement with appropriate safeguards for that transfer. The AI provider keeps what it receives for no more than 30 days and does not use it to train its models.
Cases with AI for talent. A company can choose a case in which the AI only helps participants. Then everything above that helps you happens (the response, the check, the summary and your feedback), but the company sees nothing of the AI assessment: no advice, no scores, no passages, and not the AI's response to your first version either. It reads your solution itself and chooses without advice. It does read your summary, which you approved yourself. The case states in advance which type it is.
Cases without AI. A company can choose a case without AI. Then none of the above happens: nothing from the case or your solution is sent to Anthropic or any other AI service, no AI assesses or summarises, and you receive no AI feedback. The company reads your solution itself. The case states in advance whether it is without AI. You may still use AI yourself when creating your solution.
Based on those scores, our own system gives advice on each solution: not an exact place, but a range, for example “place 2–4”. For this, no data leaves TalentProof. That advice is not the decision: the company itself chooses which solutions it reads — as many as the case has winners, one to five — from the solutions the AI ranks highest (three to eight) or from outside them, and nothing is ticked in advance. The company then also decides the ranking and the winners. We keep both the advice and the company's choice. No decision is therefore taken that is based solely on automated processing. Until the results are final, the company does not see your name: it assesses your work under a number.
The AI also assesses your competencies, using the same fixed description per level (rubric) that companies use. From this, you receive written feedback after the case closes: what the AI saw in your solution, and what a next step would be. If the company does not assess your competencies, you will also see the AI's estimate per competency after the results — only you, and it does not go on your profile. Companies never see the AI's scores. However, when the company assesses each competency, it is shown the parts of your solution that the AI considered relevant to it, without a score or judgement: the company decides for itself whether they are strong or weak. We also use the AI's scores to check whether the AI and companies see the same thing, and as a signal for the participation guarantee we give the company; that guarantee counts the solutions that clear the bar (they make it plausible that they meet at least half of the success criteria). Where necessary, a person at TalentProof decides on this, and it has no consequences for you. If a company reports that your solution on its shortlist does not clear the bar, a person at TalentProof decides as well, not the AI; if it does not clear the bar, it drops off the shortlist, and you will see why after the results. Anything you do not see in the app, you can always request from us. What goes on your profile is the rating of a company that read your solution: before the results are final, it uses that rubric to give a level, or a range of two levels, for each competency, without seeing your name.
While you are still working on your solution, the AI responds to your work once. After you have locked the first version of your solution, that version is sent with the case (without your name) to Anthropic. The AI says what works, what is missing or does not quite fit, and what questions a decision-maker would ask, without a score or judgement. In the step “The AI's take”, you write what you do with it. We store the response with your solution, and the company reads it alongside what you did with it. If the AI cannot respond at that moment, that step may be left empty.
Just before you submit, you can have the AI check whether your solution covers everything the company asks for: what you are asked to submit and each success criterion. You can do this twice. For this, your solution (without your name and without the summary) is sent with the case to Anthropic. For each point, the AI says whether it is covered, not how well. That result is for you only: the company does not see it, and it does not count in the assessment. We store the latest result with your solution.
At the end, you write the core of your advice in one or two sentences; this counts in the assessment. The AI then drafts a summary of your solution once, in two short parts: what you propose and why it works. For this, your solution (without your name) is sent to Anthropic once more. You edit the summary and confirm that it is correct, or you leave it empty. The company sees it next to your core, with a note that the AI drafted it and you approved it. The summary does not count in the assessment, but the company chooses its shortlist based on your core and your summary.
If you disagree with your assessment, you can ask on your own submission for a person to look at it. Someone from our team then reads your solution themselves and you receive the outcome. That is different from running the model again: then the same assessor would look at the same work once more.
The AI provider processes your solution outside the European Economic Area, under a data processing agreement with appropriate safeguards for that transfer. The AI provider keeps what it receives for no more than 30 days and does not use it to train its models.
Cases with AI for talent. A company can choose a case in which the AI only helps participants. Then everything above that helps you happens (the response, the check, the summary and your feedback), but the company sees nothing of the AI assessment: no advice, no scores, no passages, and not the AI's response to your first version either. It reads your solution itself and chooses without advice. It does read your summary, which you approved yourself. The case states in advance which type it is.
Cases without AI. A company can choose a case without AI. Then none of the above happens: nothing from the case or your solution is sent to Anthropic or any other AI service, no AI assesses or summarises, and you receive no AI feedback. The company reads your solution itself. The case states in advance whether it is without AI. You may still use AI yourself when creating your solution.
5. Retention periods
How long we keep what, and why. A daily run deletes whatever has passed its retention period.
| What | How long | Why |
|---|---|---|
| Your account and profile | As long as your account exists. If you don't log in for 24 months, we send you an email; if you then don't log in within 30 days, we delete your account. | So you can take part, and so we don't keep data about people who no longer use the platform. |
| Your solutions, their AI assessment (including exactly what the AI received) and the companies' ratings | As long as your account exists. | They make up your profile and the proof of your work. The EU AI Act requires the assessment to be recorded. |
| Changes you save after submitting but don't submit again | Until the case closes. | Until the case closes you can edit your solution. Anything you didn't submit again doesn't count and is no longer needed after that. |
| The recruitment record between a company and you, and what you are open to at a company | As long as you both have an account. If either of you deletes their account, the record goes with it. | To establish whether an appointment came about through TalentProof (terms, article 8). |
| Company accounts and cases | As long as the company is a customer; earlier on request, except what must be kept for tax purposes. If a company deletes a case after the results, a short version (title, sector, key question and description, with or without the company name) stays with the work participants put on their profile. | To be able to deliver the case and the results. |
| Photos we add to a case | As long as the case exists. We delete an uploaded photo that hasn't been attached to any case for 7 days. | To display the case. The photo is stored on our own server, so your browser doesn't contact other websites. |
| The logo a company puts on its profile | As long as it is on the profile. We delete a replaced or removed logo straight away; if the company account is deleted, the logo goes with it within 7 days. | To display the company. The logo is stored on our own server, so your browser doesn't contact other websites. |
| Questions for a company in a case's Q&A round, and the answers | As long as the case exists; if the company deletes the case after the results, they go with it. If you delete your account, your questions that hadn't been answered yet go with it; an answered question stays, without your name. | So that everyone working on the case has the same answers. |
| Invoices and payments | 7 years after the end of the year. | Statutory obligation to keep accounting records. |
| Payout details (name, address, date of birth, IBAN) if you received prize money or a share | 7 years after the end of the year of payout, even if you delete your account. | The Dutch Tax Administration (Belastingdienst) can request these details for payments to third parties. |
| Notifications in the app | 12 months after you read them. | To keep you informed; no longer needed after that. |
| Cancelled or failed payment attempts | 12 months. | To be able to trace a payment. |
| Counters against login abuse (email address and IP address) | 7 days. | Protection against password guessing. |
| Visitor statistics (only counts per day, page, source, device and browser) | 26 months. | To see how the site is used and where visitors come from. Without IP address or cookie, not traceable to a person. |
| Backups | 14 days; anything you deleted then also disappears from the backups. | To be able to restore the platform after an outage. |
6. Sharing with third parties
We share data with processors that help us provide the platform, each under a data processing agreement:
If you take part in a case, the company behind that case sees, after the case closes, your core, your summary and the AI's advice on your solution; your name and the solution score only once the results are final, and then also what you currently do and what you are open to (including at this company), if you have filled this in, and the recruitment record. The company sees your full solution if it is on its shortlist, or if the company buys it. The company does not see the AI's competency scores; it does see, when assessing, the parts of your solution that the AI found relevant for each competency. If you ask a question in the Q&A round, the company sees that question without your name.
If you log in with Google or Microsoft, we receive from that service your name, your email address and a fixed number by which that service recognises you. We use that number to link your login to your account, so that nobody else can get in with your email address. What Google and Microsoft process themselves is set out in their own privacy statements.
If you use your own AI (ChatGPT, Copilot, Claude, Gemini, etc.) when creating your solution, what you enter there is governed by the terms and privacy policy of that service, not by this statement: we do not see it and have no control over it.
Beyond that, we only share data if you choose to: if you make your profile public, your name, pitch, results, your level and companies' ratings of your competencies are visible to everyone. We calculate that level (Participant, Winner or Top talent) from what companies did in competitions with results: how often you submitted and were among the winners, and at how many companies. No AI is involved in this, and you can see on your profile how it is made up. After the results, you can also put the core and the summary of each solution on your profile, if the company allows it, together with a short version of the case. Never your whole solution, and you can always remove them again. We do not sell your data.
- Anthropic — assessment of solutions by AI (see section 4); not for cases without AI.
- Mollie — payments. We do not see or store card details.
- Our hosting providers — for the server the platform runs on and for the mail server we use to send email.
If you take part in a case, the company behind that case sees, after the case closes, your core, your summary and the AI's advice on your solution; your name and the solution score only once the results are final, and then also what you currently do and what you are open to (including at this company), if you have filled this in, and the recruitment record. The company sees your full solution if it is on its shortlist, or if the company buys it. The company does not see the AI's competency scores; it does see, when assessing, the parts of your solution that the AI found relevant for each competency. If you ask a question in the Q&A round, the company sees that question without your name.
If you log in with Google or Microsoft, we receive from that service your name, your email address and a fixed number by which that service recognises you. We use that number to link your login to your account, so that nobody else can get in with your email address. What Google and Microsoft process themselves is set out in their own privacy statements.
If you use your own AI (ChatGPT, Copilot, Claude, Gemini, etc.) when creating your solution, what you enter there is governed by the terms and privacy policy of that service, not by this statement: we do not see it and have no control over it.
Beyond that, we only share data if you choose to: if you make your profile public, your name, pitch, results, your level and companies' ratings of your competencies are visible to everyone. We calculate that level (Participant, Winner or Top talent) from what companies did in competitions with results: how often you submitted and were among the winners, and at how many companies. No AI is involved in this, and you can see on your profile how it is made up. After the results, you can also put the core and the summary of each solution on your profile, if the company allows it, together with a short version of the case. Never your whole solution, and you can always remove them again. We do not sell your data.
7. Your rights
You have the right of access, rectification, erasure, restriction, objection and data portability. You can also lodge a complaint with the Dutch Data Protection Authority (Autoriteit Persoonsgegevens). You can delete your account yourself, under Settings: you will receive a confirmation link by email. For your other rights, contact us.
8. Cookies and visitor statistics
We only use a necessary cookie to keep you logged in, and a functional cookie that remembers for a year which language you read the site in (Dutch or English). No tracking or advertising cookies.
To see how the site is used, we count page views ourselves, on our own server and without involving any other party. For each view, we add one to the count for: the day, the page, where you came from, the type of device and the browser. We first remove numbers and names from the address of a page, such as those of a case or a profile. We do not store any IP address, do not set a cookie and do not store individual visits: what remains are counts that cannot be traced back to you. Visits to admin pages and visits by bots are not counted.
To see how the site is used, we count page views ourselves, on our own server and without involving any other party. For each view, we add one to the count for: the day, the page, where you came from, the type of device and the browser. We first remove numbers and names from the address of a page, such as those of a case or a profile. We do not store any IP address, do not set a cookie and do not store individual visits: what remains are counts that cannot be traced back to you. Visits to admin pages and visits by bots are not counted.
9. Contact
For questions about this statement or your data: email info@talentproof.nl. TalentProof is a trade name of Solutit, Fruitweidestraat 91, 8043 XD Zwolle, registered with the Netherlands Chamber of Commerce (KvK) under number 50836552.
See also our terms of use.