Legal
Privacy policy
What we collect, why we are allowed to, who else touches it, where it goes and how long we keep it. It is written to be read, not to be survived.
Updated
At a glance
- The AI
- What you ask the AI to work on, such as your resume and a job text, is sent to Scaleway, a French company, and processed in its data centres in Paris. Section 4 says what that means.
- Where it runs
- Your saved data sits in a database in Europe. Our servers run in the USA.
- What we do not do
- We do not sell your data, use it for advertising, or decide anything about you automatically.
- Cookies
- One sign-in cookie. No tracking cookies, so no consent banner.
- Your rights
- Download your data or delete your account in Settings, under Data, or write to us.
1. Who we are
Responsa is run by Copenhagen Workshops, a sole proprietorship registered in Denmark (CVR 46388526). We are the data controller for the personal data described here.
Contact for anything in this policy (access, deletion, corrections, complaints): hello@responsa.dk.
If you are in the EU or EEA you can also complain to the Danish Data Protection Agency, Datatilsynet, or to your own supervisory authority.
2. What we collect
Resume content. The resume you paste or upload, and the structured data we extract from it: your name, contact details (e-mail, phone, location), work history, education, skills, links, and, only if you add one, a profile photo. Whatever you put in your resume, we process; if you include something sensitive (health, religion, trade-union membership and so on), it travels with the rest of the document.
Job posting content. The job text you paste, or a job link you ask us to read for you.
Account details. Your e-mail address, and your name if you sign in with Google or LinkedIn. There are no passwords: you sign in with a one-time code sent to your e-mail, or with Google or LinkedIn. An account is needed to use the AI features and to see the jobs.
Saved documents. The master resumes you save, the tailored resumes and cover letters you keep, the job postings you save with them and your notes on each application, including any photo embedded in them.
Your search profile. If you open My Career Paths: a short summary our AI reads from your saved resumes and the job titles you applied for. It holds the roles you could apply for, each with the words of your resume that show it, your employers and their industries, how many years you worked in each, your strongest skills, what your experience shows in short phrases (such as “executive dashboards”), your field of research if you have one, and your level. It is used to draw your career map and to find jobs for you in Jobs for You. It is read again when your resumes or the jobs you applied for change, it is deleted when you delete your last resume, and you can see all of it in Download my data.
Your career map. If you open My Career Paths: the positions your saved resumes are written for, the positions each could lead to, and one sentence for each on why it fits you, which our AI writes from your search profile and your saved resumes. It is used only to show you that map, and it is in Download my data.
Your job searches. If you use Find Jobs: the words you type into the search line, and the positions our AI reads from what you write in the Jobs for You line. They are used only to learn which positions you look for.
Your Jobs for You picks. The thirty jobs our AI picks for you from those that fit your filters, each with a one-line reason and the scores it gave on the way, kept so that when you come back only the new jobs are scored again.
Your answers in the Jobs for You conversation. The first time, our AI asks a few questions about the work you want. We keep only what you agree on there: a one-line summary of the work you want, the positions to look for and the ones not to show, the countries, towns and regions, languages, levels and kinds of workplace you choose, and whether to show student jobs. We never keep the conversation itself: it stays in your browser until it is over. What you write in it, with the search profile read from your resumes, is sent to our AI provider (section 4) to answer you.
Technical data. Your IP address, used for rate limiting so a script cannot exhaust the service (section 7 says how briefly), and the standard logs of our host.
Usage events. Which features you use and whether they worked, for example that a PDF was downloaded. They carry no resume or job text. Section 9 explains the two kinds and how they differ.
Pro waitlist. If you leave your e-mail on the Pro card of the landing page, we keep that address and the date, nothing else, and write to it once, on the day Pro opens. It is not tied to an account.
Payments. If you buy credits or the unlimited plan, you pay on Stripe’s own page, run by Stripe Payments Europe, Ltd. Your card details go to Stripe and never reach us. Stripe receives your account e-mail address, an account number of ours that ties the payment to your account, and the payment itself. We keep Stripe’s customer, subscription and payment numbers and what they tell us: your plan, when its paid period ends and whether it is cancelling, and your credit balance with the credits you bought. If you delete your account, your customer record and receipts stay in Stripe, because the bookkeeping rules require us to keep them. If you withdraw from a purchase, we keep a record of it: the purchase, the day you withdrew, the name and e-mail you gave and the amounts paid, kept and refunded. It is part of our books, so it stays when you delete your account, no longer tied to it.
E-mails about a purchase. Consumer law asks us to confirm a purchase in writing, and to confirm a withdrawal. So when you buy we e-mail you the confirmation of the contract, with the last day to withdraw and the withdrawal form, and when you withdraw we e-mail the receipt. They go to the address you gave Stripe, or on the withdrawal form, through Resend (section 5). We keep which purchases have had their confirmation, by Stripe’s payment number only, no address.
3. Why we use it, and our legal basis
| What | Why | Legal basis (GDPR Art. 6) |
|---|---|---|
| Resume and job text | To produce your tailored resume, cover letter and fit assessment. This is the service | 6(1)(b): performing the contract you asked for |
| Account details | To let you sign in and find your saved work | 6(1)(b) |
| Saved documents | So you do not re-upload your resume every visit | 6(1)(b) |
| IP address | Rate limiting and abuse prevention | 6(1)(f): our legitimate interest in keeping the service available |
| Server logs | Finding out why something failed | 6(1)(f) |
| Product analytics | To see which features are used and what breaks | 6(1)(f). Section 9 explains why no consent banner is involved |
| Usage history of a signed-in account | To see where people get stuck, to answer support questions with your real history in front of us, and to decide what to fix | 6(1)(f). You can object at any time (sections 8 and 9) |
| Operator access to saved documents | To look into a problem you report and to check the quality of what the product produced | 6(1)(b) for support you ask for, 6(1)(f) for quality checks (section 9) |
| Jobs for You picks and conversation answers | To pick the jobs that suit you and say why, and to set your filters from your answers | 6(1)(b): you asked for it by opening the view |
| Credits: your balance and each change to it | To count what you can still run and show you your balance | 6(1)(b) |
| E-mails we send you: your sign-in link, the confirmation of a purchase, the receipt of a withdrawal | To let you in, and to confirm in writing what you bought and what you withdrew from, as consumer law asks | 6(1)(b) for the sign-in link; 6(1)(c): our legal duty, for the confirmation and the receipt |
| Pro waitlist e-mail | To write to you once, when Pro opens | 6(1)(a): your consent, given by sending the form. Withdraw it by writing to us |
We do not sell your data and we do not use it for advertising. We keep a usage history for each signed-in account (section 9); it is used only to run and improve the product and to help you, and it is not shared with anyone.
4. AI processing: what leaves our servers, and where it goes
This is the part most policies bury, so it gets its own section.
Our AI runs at Scaleway SAS, a French company, in its data centres in Paris. When you ask the AI to tailor a resume, write a letter or a follow-up message, build a resume in the chat, read a resume you upload or a job you paste, or assess a fit, we send Scaleway the text it needs: your resume and the job posting. Not a summary; the whole document, on each round. There is no way to provide these features without doing that. Searching and filtering jobs and tracking your applications do not send your resume to the AI.
What that means for your data:
- The model is an open model published by DeepSeek under the MIT licence. Scaleway runs its own copy on its own servers. DeepSeek, the company, does not receive your text, and Scaleway says the makers of the models cannot access what its customers send.
- Scaleway is our processor, under its Data Processing Agreement. It says it does not collect, read, reuse or analyse the content of what we send or of the answers, and that it does not use it to train or improve the models.
- Scaleway does not store the content in normal operation. One exception: if a request harms the service, for example by causing errors, Scaleway may keep and read that customer's requests, text included, for up to two weeks to find and fix the fault. Your resume could be among them.
- Scaleway processes in Paris today and has said it may add other sites in Europe. If where your text is processed changes, we change this section the same day.
- Scaleway lists its own sub-processors on its website: operators of the data centres it rents space in. Some are headquartered in the USA; the sites it lists for them are in the EU.
- We send the text to produce your result and for nothing else, and we do not keep it ourselves beyond the request (section 7).
When Scaleway is busy or fails, the request goes first to a second model at Scaleway, in Paris too. Only if that cannot answer either does it go, once, to OpenRouter, Inc., a gateway in the USA. We tell OpenRouter to pass it only to a host that has promised to keep no copy and not to train on it, and never to DeepSeek's own service or to the other Chinese companies that run this model. The host that answers may sit outside the EU. On an ordinary day nothing goes to OpenRouter. No one else is asked: if OpenRouter cannot answer either, the request fails and you can try again.
Other things in this section:
- Company research. To write a short note on the company you apply to, we search the web with Tavily (USA) for the company's name only, never your resume or the job text, and send the pages found to the AI to write the note. The company's name is read from the job posting by the AI.
- Jobs for You. To pick your thirty, we send the AI your search profile (section 2): the roles you could do, your industries and years, your strongest skills, your level. Never your resume's text, name or contact details. With it go short codes of the jobs that fit your filters (title, employer, languages, level, terms) and, for the fifty closest, the sections of those postings about the role and its requirements.
- Your photo is never sent to the AI. Only the code that builds your PDF or Word file handles it.
- No decision with legal or similarly significant effect is made about you. The score and fit assessment are suggestions for a document you control; nothing is decided about you automatically in the sense of GDPR Article 22. You accept, edit or ignore every suggestion.
5. Who processes data for us
We share the minimum each of these needs.
| Processor | What it does | Where | Transfer basis |
|---|---|---|---|
| Scaleway SAS | The AI: receives your resume and job text and writes the result (section 4) | France (Paris) | Within the EU. Scaleway is our processor under its Data Processing Agreement |
| OpenRouter, Inc. | Only when Scaleway cannot answer (section 4): receives the same text as Scaleway, your resume and the job posting, and passes it to a host that has promised to keep no copy and not to train on it. Never DeepSeek’s own service or other Chinese hosts | USA; the host it picks may be elsewhere | Standard Contractual Clauses, in OpenRouter’s Data Processing Agreement |
| Tavily | Web search for a company by its name; never your resume or the job text | USA | No personal data of yours is sent |
| Supabase | Sign-in, and the database that holds your saved documents. Its sign-in e-mails are sent through Resend, below | Europe | Stored in Europe |
| Resend (Plus Five Five, Inc.) | Sends our e-mails: your sign-in link, the confirmation of a purchase and the receipt of a withdrawal. Receives your e-mail address and the text of the message, and keeps a log of what it sent. Our e-mails are sent from its Ireland region | USA (the company); sending runs in Ireland | Standard Contractual Clauses, in Resend’s Data Processing Agreement |
| Stripe Payments Europe, Ltd. | Takes payment for credits and the unlimited plan, on its own page: receives your account e-mail address and the payment. Your card details never reach us | Ireland | Within the EU. Stripe’s own terms apply to any transfer outside it |
| Vercel, Inc. | Hosting, and the servers that run the app, build your PDF and Word files and pass your text on to the AI provider | USA: the servers run in Washington, D.C. | EU–US Data Privacy Framework, or Standard Contractual Clauses |
| Upstash | Rate-limiting store; holds your IP address for 60 seconds | Frankfurt, Germany | Standard Contractual Clauses (the company is American) |
| PostHog | Product analytics; no resume or job text | EU cloud (eu.posthog.com) | Within the EU |
| Jina AI GmbH | Reads a job posting from its web address when you paste a link, open a job in Find Jobs that came without its text, or ask for the full ad on a public job page. Only the posting’s address is sent, nothing of yours | Germany | Within the EU |
| People Data Labs, Inc. | Only when you import a resume from a LinkedIn link: receives the web address of the profile you paste, and returns the work history, schools and skills held for it. Nothing else of yours is sent, and we use none of its other fields (contact details, birth date) | USA | EU–US Data Privacy Framework (the company is certified), or Standard Contractual Clauses |
| Only if you sign in with Google: supplies your e-mail and name; Google sees that you signed in | USA | EU–US Data Privacy Framework, or Standard Contractual Clauses | |
| Only if you sign in with LinkedIn: supplies your e-mail and name; LinkedIn sees that you signed in | USA and Ireland | EU–US Data Privacy Framework, or Standard Contractual Clauses | |
| DuckDuckGo | Supplies the small logo shown next to each company. Our server asks for it, not your browser: DuckDuckGo sees our request for a company’s domain, never your IP address, and nothing you type or upload | USA | No personal data is sent |
International transfers
Where a processor sits outside the EEA, the transfer runs on the European Commission's Standard Contractual Clauses or the EU–US Data Privacy Framework. The AI is processed inside the EU: your resume and job text are handled by Scaleway in France (section 4), except on the rare request Scaleway cannot answer, which goes to OpenRouter in the USA on the same Standard Contractual Clauses. Our servers in the USA, Vercel's, pass that text on to the AI without keeping it.
Open data behind the Region filter
The Region filter in Find Jobs joins towns to regions by us, with data from GeoNames (geonames.org) and from Klimadatastyrelsen (DAWA: Danish postal districts, municipalities and town names, retrieved 30 September 2026). Both are licensed under CC BY 4.0. We changed the data: the towns were matched to regions and reduced to the lists the filter needs. These are lists of places stored in the app; nothing about you is sent to either source.
6. What is stored in your browser
- A sign-in cookie, if you are signed in. It is strictly necessary: without it you would be signed out on every page.
- Small settings in your browser's local storage, such as your light or dark theme and that you closed the cookie note.
We use no third-party tracking cookies, no advertising cookies and no cross-site identifiers. Nothing we set is optional, so there is no cookie banner to click through; section 9 explains.
7. How long we keep it
| Data | Kept |
|---|---|
| Resume and job text sent to the AI for a request | Not kept by us beyond the request. What you save of the result is kept as saved documents, below |
| Saved documents (master and tailored resumes, letters, saved postings) | Until you delete them, or delete your account |
| Search profile | Until you delete your last resume, all your documents or your account; replaced when your resumes or the jobs you applied for change |
| Career map | Until you delete your last resume, all your documents or your account; replaced when you ask for a new map |
| Job searches on Find Jobs | Until you delete all your documents or your account |
| Jobs for You picks | Until you delete all your documents or your account; replaced when you open Jobs for You on another day |
| Answers in the Jobs for You conversation | Until you delete all your documents or your account. The conversation itself is not stored |
| Account details | Until you delete your account |
| Pro waitlist e-mail | Until Pro opens and the one e-mail is sent, or until you ask us to delete it |
| IP address in the rate limiter | 60 seconds, the length of the rate-limit window |
| Server logs | By our host, for the short period its plan allows. They hold no resume or job text |
| Product analytics events (PostHog) | They carry no identifier, so they cannot be tied back to you, and no resume or job text |
| Usage history (our own records, section 9) | 90 days as records; after that only daily counts (how many times a page was opened or a button pressed that day), until you delete your account. Deleting your account deletes both |
| Credit balance and its history | Until you delete your account. A count of AI runs stays without your account after that, with nothing that points to you |
| Record of a withdrawal from a purchase | Five years after the end of the financial year it happened in, as the Danish Bookkeeping Act requires; deleting your account does not delete it |
| Record of operator access to your documents | Kept, so every opening can be accounted for |
| Database backups | Our database host may keep short-lived backups. Data you delete can survive in one until it expires |
8. Your rights
Under the GDPR you can:
- Access the data we hold about you.
- Correct it, in the app or by writing to us.
- Delete it. You can delete saved resumes one by one, and your whole account in Settings, under Data. The account is deleted 30 days after you ask, and you can withdraw the request before then.
- Export it: Settings, under Data, downloads a copy of your data in a portable file.
- Object to processing based on legitimate interest (section 3), including analytics.
- Withdraw consent where we rely on it, as for the Pro waitlist.
- Complain to Datatilsynet, or to your own supervisory authority.
Write to hello@responsa.dk for any of these. We answer within one month, as the GDPR requires.
9. Analytics: the honest version
There are two different things here, and mixing them up is how most policies get this wrong.
Product analytics in your browser. We record which features you use and whether they worked: feature names and outcomes, nothing more. Our analytics tool runs with automatic page and click capture switched off and with no identifier stored in your browser, so an event cannot be linked back to you or to your earlier visits, and nothing survives after you close the tab. This is why the service has no cookie banner: there is nothing stored on your device to ask permission for.
Our own usage records (since 22 September 2026). Separately, our server records what happens in the product: which page was opened, which button was pressed (by its name, such as “Download PDF”), which feature was used and whether it worked, and when a tab was left. The tab holds a random visit id in memory only, and if you are signed in, our server attaches your account to these records from your sign-in. What is never recorded: anything you type, and anything written in your resume or cover letter; a click inside the document is recorded without the text clicked. The records are kept for 90 days and then reduced to daily counts. Both live in our own database; they are not sent to any analytics company. You can object at any time and we will stop recording your account (section 8).
Who can see your saved documents. The service is run by one person, who can open a signed-in user's saved resumes, job postings and cover letters in an internal admin tool, to look into a problem you report and to check the quality of what the product produced. That tool is protected by a second factor (a code from an authenticator app), and every opening of a document is written to a log.
Server records. Our servers keep short technical records of things that happened: a request came in, it was rate-limited, an export failed. We need them to run and debug the service. They do not contain your resume or job text.
If we ever add analytics that store an identifier on your device, a consent banner comes with it, and this section will say so first.
10. Security
Access to your saved resumes is enforced in the database: a row can be read only by the account that created it (row-level security), not merely hidden by the app. There are no passwords to leak: sign-in is a one-time code or Google or LinkedIn. Traffic is encrypted in transit. Our keys live in the hosting platform's secret storage, never in the code sent to your browser.
11. Children
Responsa is not meant for children. You must be at least 16 to use it. If you believe a child has given us personal data, write to us and we will delete it.
12. Changes to this policy
If what we collect or who handles it changes in a way that matters, such as the AI provider, we will say so by changing the date at the top and, when it matters to you, with a notice in the app or by e-mail to registered users. Small wording fixes just get a new date.
13. Contact
Copenhagen Workshops, CVR 46388526. hello@responsa.dk