What the CV anonymiser removes
Recruiters anonymise CVs for two different reasons. The commercial reason is to stop a client contacting a candidate directly before an introduction is agreed. The fairness reason is to let a hiring manager judge skills and experience without seeing a name, an age or a nationality. This tool helps with both by finding the details that most often identify a person and replacing each one with a labelled placeholder such as [EMAIL] or [NAME], so the reader can see that something was removed.
- Names: the candidate’s name on the opening lines, every later mention of it (including first name only and possessives), names you type in, labelled names and referees listed under a References heading.
- Contact details: email addresses (including “name [at] domain” spellings), phone numbers in UK, US and international formats, web addresses, LinkedIn and GitHub profile links and social media handles.
- Postal addresses: street addresses with a house number, UK postcodes, US state and ZIP codes, Canadian postcodes and lines labelled “Address”.
- Date of birth and age: labelled dates of birth, “born on” dates, “Age: 34” and “34 years old”.
- Nationality and right to work: nationality, citizenship, visa, passport, work permit and sponsorship lines.
- Other personal details: labelled gender, marital status, religion, health, dependants and identity numbers, which an employer rarely needs at shortlist stage.
How to anonymise a CV before sending it to a client
- Paste the CV text or upload the candidate’s PDF or DOCX. Check that the text reads in the right order; two-column PDFs sometimes come out interleaved.
- Type the candidate’s name if you know it. The detector then redacts every occurrence, including a first name used on its own in the profile.
- Select Find personal details and read the highlighted original. Each colour is one category.
- Switch off anything that is not personal data, such as an employer that happens to share a surname. Switch categories off as a whole if your client policy allows them, for example keeping the town but removing the street.
- Read the redacted text from top to bottom, then copy it or download it as a .txt file.
What heuristic redaction misses
Pattern matching is fast and predictable, but it is not understanding. It can miss a nickname, a name written in lower case, an address without a house number or postcode, a phone number split across lines, or a personal detail mentioned in passing (“relocating to be near my partner in Leeds”). It does not see text inside images, logos or photos, and plain text cannot carry over hidden document properties such as the author field of a Word file. That is why every result is shown for review and nothing is redacted silently.
The opposite error also happens: a company called Morgan & Co will be flagged if the candidate is called Taylor Morgan, and some product codes look like postcodes. Switching a detection off keeps that text unchanged in the output.
Contact removal is not full anonymisation
A CV with no name, email or phone number can still identify someone. A unique job title at a named employer, an award, a small university department or a published paper may be enough for a client who knows the market. For a genuinely blind shortlist, also generalise employers (“a FTSE 250 retailer”), institutions and exact dates, and remove photos. Data protection law in the UK and EU expects you to share only the personal data a purpose needs, so a short, deliberate redaction policy per client is worth agreeing in advance.
Why section headings are never redacted
Headings like References, Experience, Education and Personal details are part of the document’s structure. A redaction tool that turns the References heading into a name placeholder makes the CV harder to follow and leaves the referees themselves exposed. This anonymiser keeps a list of common CV headings and never changes a line that is only a heading, while still checking the lines underneath it.
Where your text goes
The detection runs in your browser. Pasted text is not uploaded. If you choose a file, it is sent once to our file reader, processed in memory to extract the text and discarded; nothing is stored or written to logs, and the page’s analytics only count that the anonymiser was used. Use synthetic examples if your organisation does not allow candidate data in web tools.
From redacted text to a client-ready submission
Plain text is fine for a quick check, but clients expect your agency’s format. CVPitch extracts the candidate’s facts into your branded template, shows the source passage beside every fact, removes the contact details you select from the Word file, the PDF and their hidden metadata, and ties approval to the exact version you send. Compare monthly plans or start with 10 free CVs.