Privacy
Your data, plainly
Who runs Evident Career Studio and how to contact us
David Hall, trading as Evident Career Studio, runs this site and is the data controller for the personal data described on this page. Postal address: 7 Raynsford Road, Ware, Herts, SG12 7AJ. To ask about your data or make a privacy request, email evidentcareerhelp@gmail.com or use the on-site form.
Make a privacy requestThe form records your request as a case and gives you a reference and a private key. Keep the key safe, you need both to check the status or delete the case on the case page, and we can't recover a lost key. Giving an email address is optional. Case status updates are shown on the case page and are not emailed automatically.
- Your files
- PDF, Word and text files are read inside your browser. The file itself is never uploaded or kept. Relevant extracted or entered text is sent to AI providers when you ask for questions, reads, drafts, matching or the final build, as described below.
- Where your CV and answers live
- The free scan checks run in your browser. If you ask for AI help, the relevant text is sent to our AI providers to produce it. Your work is saved in this browser. During the final CV build, working material is held encrypted on our server. The finished CV is held encrypted for up to 30 minutes after completion, then deleted, or sooner once received by your browser. See the privacy page for provider handling and optional sharing. Your CV text, job text, answers, AI-suggested and approved wording, organisation corrections, application pack choices and saved complaint references and keys are saved in this device's browser storage. Clearing browser data or using the button below removes those browser copies, not a submitted complaint, an optional shared run, or text handled by an AI provider under its own terms.
- Your analysis and questions (Anthropic)
- When you ask for questions, and when Evident reads the layout of your CV, your CV text and the job text are sent from our server to Anthropic's Claude AI, which produces the analysis and your questions. Anthropic may keep that text for a limited period under its own commercial terms and policies, which we don't control, see anthropic.com/legal. This app's own code doesn't save that text on our server, doesn't write it to our logs and doesn't put it in analytics.
- Follow-ups and draft wording (OpenAI)
- When you answer a question and ask for a follow-up or draft wording, the relevant text, which can include the question, your answer and short extracts of your CV and the job text, is sent from our server to an AI model from OpenAI through Lovable's AI service, on its zero data retention setting. This app's own code doesn't save that text on our server, doesn't write it to our logs and doesn't put it in analytics. Lovable and OpenAI handle it under their own terms, see lovable.dev/privacy and openai.com/policies.
- Matching your CV to the job (OpenAI)
- When you press “Build my CV”, the job's requirement lines and the lines of your CV (not your answers or contact details) are sent from our server to an AI model from OpenAI through Lovable's AI service, on its zero data retention setting. It only points out which of your CV lines mean the same as each requirement, so a skill written in different words is not missed. It writes no text for your CV. Our server doesn't save or log that text.
- Writing an optional cover letter (Anthropic)
- Only if you press “Write my cover letter”: your CV text, the job advert, your approved answers and your finished CV are sent from our server to Anthropic's Claude AI, which writes the letter. Anthropic may keep that text for a limited period under its own commercial terms and policies, which we don't control, see anthropic.com/legal. We don't save the letter on our server, log it or put it in analytics. It stays in this browser until you clear your session. We count how many letters each access code writes in a day (a scrambled code and a time, no text), deleted after 7 days.
- Writing your final CV (Anthropic)
- When you press “Build my CV”, your confirmed CV text, the job advert text and the answers you approved (never ones you marked private) are sent from our server to Anthropic's Claude AI, which writes the CV. This is the same provider as for your analysis and questions. Anthropic may keep that text for a limited period under its own commercial terms and policies, which we don't control, see anthropic.com/legal. Our server never logs it or puts it in analytics. To protect you if your connection drops, your CV details and the finished CV are held encrypted on our server while it is being written and for up to 30 minutes after, then deleted. They are deleted sooner once your browser has received the CV. We do not otherwise keep CVs, adverts or answers, unless you choose to save a run for the pilot review (see below). What is held: your CV text, advert text and approved answers (encrypted, only while the CV is being written), then the finished CV (encrypted), plus a scrambled code for the build, its status and times. It can only be opened with a key kept in this browser, so it can be recovered only on this device and browser. A build that fails keeps only an error code and reference, never your text. The finished CV then stays in your browser.
- Complaints you submit
- If you submit a complaint, we store on our server: the category, your description, your email if you give one, basic context (which step and item, and your browser type, not the text), and a redacted excerpt only if you tick the box after seeing exactly what will be sent. We also keep a scrambled (hashed) form of your connection's IP address and of your private key, to stop abuse and so only you can open the case. We never store the raw IP address or the key itself.
- Who can see a complaint
- Only someone with both the case reference and the private key, and the site owner through a protected, signed-in owner account. Other visitors cannot list or search cases.
- How long complaints are kept
- A case is deleted automatically 180 days after it was last changed (submitted or updated). The one exception: a case marked “In review” is kept until it leaves review, then the 180 days apply. You can delete your case sooner at any time on the case page. The clean-up runs once a day, so deletion can take up to a day longer.
- How long fair-use records are kept
- Records used for rate limits, a scrambled IP address, the type of request and the time, are deleted automatically after 7 days by the same daily clean-up.
- Where the server is
- Our database runs on Lovable Cloud. Its connection address points to Amazon Web Services' eu-west-1 region (Ireland). We haven't independently confirmed where backups or the AI service process data, so we don't make a wider claim about where data is kept.
- AI, not people
- Questions and suggested wording are generated by AI. No person reviews your CV or answers.
- No verification of you
- We never verify employment, attendance, qualifications or dates. Organisation names are picked out of what you wrote, and you can correct them. Links to official registers only help you check an organisation exists.
- Speech input
- Dictation uses your browser's built-in speech recognition, which may be processed by your browser's provider (for example Google or Apple).
- Free CV scan AI read (Anthropic)
- The nine checks and the timeline on the scan page run in your browser and nothing is uploaded. Only if you press Get my AI read, your CV text (the first 12,000 characters) and the timeline notes are sent from our server to Anthropic's Claude AI for that one read. Our server does not store it, log it or use it to train anything. Anthropic may keep it for a limited period under its own terms, see anthropic.com/legal. To limit misuse we keep a scrambled connection key and a time for each read, deleted after 7 days, and a count of AI tokens used with no text.
- “Helpful?” and “Report a problem” buttons
- These send only a tag: which step you were on (for example analysis or pack), what kind of thing you rated (a question, a suggested claim, a tip or a pack line), the tag you chose (such as “Incorrect quote”), the application type you picked, and which version of our AI instructions, AI model and tips was in use. The tag record itself holds none of your CV, job text, answers, the wording you rated, your name or email. Sending a tag also creates a temporary fair-use record (a scrambled IP address, the request type and the time) that is deleted after 7 days; while it exists, it could in principle be matched to a tag by time. Tags are deleted after 90 days by the daily clean-up. The owner account can see the individual tag records and totals. Tags don't change the AI automatically. They're reviewed alongside regular checks before any change. To explain a problem in your own words, use the optional detailed report on the Help page, which you can delete.
- Fair-use limits on AI requests
- To keep the service available to everyone, each AI request records a scrambled (salted, one-way) version of your IP address, the type of request and the time, never your CV, job text or answers. These records are used only to apply hourly and daily limits.
- Payments, orders and refunds
- Payments are taken by Stripe. Your card details go straight to Stripe and never reach our server. Stripe has its own privacy notice, see stripe.com/gb/privacy.What we keep when you buy:
- A record of each order: what you bought, the amount, the time, and Stripe's reference.
- Your credits, and which application each is linked to. That is a random code, never CV text.
- A scrambled form of your email address (a one way code, not the address), so that Email me my link can find your purchase. We use the address from Stripe to send your confirmation and any link you ask for. We do not keep the address itself.
- A card fingerprint, which is Stripe's anonymous code for the card. We use it only for the one guarantee refund per customer rule.
- Any refund request: the reason you choose, your short note (up to 500 characters) and our decision.
- A small cookie that keeps your purchase on this browser. It is strictly necessary and expires after 12 months.
- Analytics
- Evident uses PostHog, in its EU region, for cookieless product analytics. On the payment choice page only, PostHog also makes a replay of how the page is used, such as scrolling, taps, pauses and error messages, so we can see where the page is unclear. Everything you type is hidden, and all text on that page is hidden, including our own prices and offer wording. Replays are never made on any page that shows your CV, a job advert or your answers, and never on Stripe's payment page. Replays are kept for 30 days and then deleted. It records which pages are viewed, the page address without any extra parts, the referring site, campaign tags in the link (utm), your browser and device type, and counts of key steps such as starting and completing scans, requesting and completing AI reads, payment confirmation shown in this browser, CV builds, download clicks with their file format, interview starts and completed feedback, rework purchases shown in this browser and refund requests. It sets no cookies and stores nothing in your browser for this. It never receives anything you type, upload or say, and never receives your CV, job advert, answers, email address or access code. It is used only to see which pages and steps work so the product can be improved, not for advertising. Our lawful basis for this is legitimate interests: we use it to fix problems in the buying flow, and we exclude all pages showing personal content. The footer badge is an image loaded from Fazier, so Fazier can see your IP address when your browser fetches it.
- Google Analytics
- We also use Google Analytics 4 for anonymous aggregate usage counts, such as page visits and key steps. Consent is denied by default and is never switched to granted, so Google Analytics sets no cookies and writes nothing to browser storage. It can still send cookieless usage requests to Google, including the page path, campaign tags (utm and gclid), browser and device information and counts of the steps you take. Other query details, titles containing personal text, CVs, job adverts, answers, emails and access codes are not sent. Private and owner pages are excluded. Google acts as a processor for this measurement service. Google receives your IP address when handling a request, but GA4 does not log or store individual IP addresses. Google signals and advertising personalisation are disabled. Our lawful basis is legitimate interests, documented in our internal review. If your browser sends Do Not Track, Google Analytics does not load.
About the free pilot (ended 18 October 2026)
The free pilot ended on 18 October 2026. These sections describe pilot records and optional sharing. Individually invited testers may still use access codes after the pilot.
- Tester feedback
- Invited testers can still fill in the feedback form after the pilot. We store choices, numbers and text typed, up to 1,000 characters per answer, the access-code label, optional study reference, whether a phone or computer was used, and submission time. We do not store IP address, name or CV text with it. Only the site owner can read it. Responses are deleted 90 days after the last change by the daily clean-up. A separate review submitted with permission to use it publicly is kept until that permission is withdrawn.
- Saving a run for the pilot review (optional)
- During the pilot, when a CV was ready, a tester could tick “Save this run for the pilot review” and press Save. Nothing was saved unless the box was ticked and Save was pressed. A saved run contained the CV text, the job advert text, the questions asked, the answers (never answers to questions marked private), the evidence list Evident read from the CV, the finished CV and a reference to the build record. It was stored in our database (Lovable Cloud, encrypted at rest by the platform), could be opened only by David Hall, the owner, was never sent to analytics and was never used to train AI. Saved runs were deleted automatically on 19 October 2026 (00:00 UK time), the day after the pilot ended.
- Optional pilot session sharing
- A tester can choose to share their full session with David Hall at the end of their CV. Nothing is shared unless the consent box is ticked and Share is pressed. A share contains CV and job text as extracted in the browser (not the files), questions, answers and follow-ups (except questions marked private), wording decisions, requirements, how the CV was built, final CV and style, quality checks, download attempts and feedback tags, with times. It is stored in Lovable Cloud, encrypted at rest by the platform, readable only by the protected owner account, never sent to analytics and never used to train AI. Shared sessions are deleted 30 days after the last update, or sooner at their recorded deletion date, and all remaining shares are scheduled for deletion after 10 November 2026. The daily clean-up can take up to a day longer. You can withdraw a share from the same browser or ask for deletion.
- Free pilot code
- During the pilot, if someone asked for a free pilot code on the home page or the CV scan page, we stored their email address and the date they signed up, and showed a personal pilot code on screen. We used the address only to limit codes to one person and to send one short feedback request after a CV was built, and nothing else: no marketing. Codes were kept only in a scrambled form that could not be turned back into the code. CVs were not linked to addresses. The list was deleted on 19 October 2026, the day after the pilot ended. The legal basis was the person's request (taking steps at their request). Questions about this list can be sent to evidentcareerhelp@gmail.com.
- Free CV count
- During the pilot, to keep it fair, we counted how many tailored CVs were built with each access code, with the date. The CV, the job advert and answers were never stored with this count. The count was deleted on 19 October 2026.
Interview practice
- People who have bought a practice interview and invited access-code holders can use it. You agree on screen before anything is sent.
- Your CV, the advert, your quick-question answers (when you come from your finished CV) and your interview answers are sent to OpenAI models through Lovable's AI service to make questions, follow-ups and feedback.
- Questions can be read aloud by Google's Gemini speech model through Lovable's AI service (question text only), and spoken answers are turned into text by your browser's own speech service (Google on Chrome, Edge and Android, Apple on Safari). You can always type instead.
- We record no audio and store no questions, answers or feedback text. A refresh ends the session.
- We keep usage counts and an estimated cost, and any rating or short comment you choose to give. Practice interview session records for paid purchases are deleted after 7 days.
Delete everything now
Removes your CV text, job text, answers, claims and saved case keys from this browser.
This clears this browser, including saved case keys. To delete a submitted complaint from our server, use the case page first.