CV builder

Write your career once. Aim it as often as you like

You are not building a document, you are building the thing every application is aimed from. Each CV is a view over that record, so a role you correct today is correct on every application showing it.

Free to start. Bring a CV you already have, or begin with nothing.

How it works

  1. Put your career in once

    Enter your roles, education, skills and projects, or upload a CV you already have as PDF, DOCX or plain text to have it read into the same structure. What comes in is shown against what is already there before it is saved.

  2. Cut a CV from it

    A CV is a named selection with its own template and language. Creating a plain one adds nothing to your career and takes nothing away: you are choosing a view, not starting another document. Starting one from a job advert is the exception — the AI writes new bullets into your real entries, badged as AI, and they stay drafts until you press "Add to my career".

  3. Say what this one shows

    Drop an entry, reorder your roles, keep three bullets on a job instead of nine. Hide everything in a section and the section itself falls off the printed page. Each is a decision about this CV alone.

  4. Print it, then come back to the record

    Open the print view and save as PDF from your browser. The next application starts from the same record, better than last time because last time's writing stayed in it.

A record you keep, not a document you re-make

Ask most people to update their CV and they open a file, scroll to the newest job and start typing — rewriting, in other words. Whatever they improve exists in that file only, and the other copies quietly go stale.

Aiming works the other way round. Your career is what gets stored: each role, qualification, project, achievement and skill an entry of its own, with its own dates and bullet points. A CV holds no words. It holds decisions — which entries appear, in what order, cut to which bullets, under which template and language — and reads the words from the record each time it is drawn.

The consequences are the design:

  • Fix a job title, a date or a clumsy bullet once and it is fixed on every CV in that language holding the entry, including ones you have not opened in a year. A translated version is a separate record, so correcting the original does not rewrite it for you.
  • A sentence you got right for one application is not stranded in it, but waiting in the record for the next.
  • Run a CV per kind of role you apply for, without that many copies of your career — how many you can create depends on your plan. Delete one and the entries behind it stay; only AI-written drafts you never added to your career go with it.

It runs the other way too: rewrite a line while looking at the CV, and the edit lands in the record.

What the record holds

A career is not only jobs and schools, so the record is not either.

  • Personal data, plus the short taglines and longer introductions a CV opens with.
  • Experience with bullet points under each role, and key achievements as entries in their own right.
  • Education and certifications, projects, publications and awards.
  • Skills grouped into categories — from a curated catalogue, or ones you add and name yourself — plus spoken languages, volunteering, interests and recommendations you ask a contact to write.

None of it obliges a CV to show it. Because each CV selects, a fuller record costs nothing — the eleven-year-old role that decides one application in twenty can wait until it earns its space.

Written once, in more than one language

Content has a source language, and a translated version lives beside the original instead of overwriting it. A CV is set to a language and renders that variant, so a Danish application and an English one stay two views of one career. The interface is English or Danish.

Getting a career in, and getting a document out

Import

An existing CV arrives as PDF, DOCX or plain text; legacy .doc is not read, so re-save it first. It is parsed into the same entries you would have typed — roles, dates, bullets, skills — then shown against what is already in your record. You merge, correct or discard, and only then is anything written.

The caveat is layout. A PDF records text in the order a program drew it, which for a two-column design is rarely the order a person reads it in, so a sidebar can come back threaded through the wrong job. Hence the review step.

Export

There is no one-click download button, and knowing that now beats hunting for it later. Exporting opens a page laid out for paper and hands the rest to your browser's print dialog, where the destination is "Save as PDF".

Templates and tailoring ask the same record different questions

A template is a layout, not a container

You pick from curated templates rather than designing one, and the template decides the page: layout, typography, where each section sits, how much has room. Because your words live in the record rather than the design, changing your mind is cheap: move a CV to another template and your selections travel with it.

Tailoring is selection with a measurement attached

Tailoring starts at creation: you paste the advert when you make the CV, not afterwards. Its requirements come back in the advert's own wording, along with how heavily it leans on hard skills, context and recency, seniority, and soft skills or domain. That reading is then frozen, and the score itself is arithmetic — recomputed on every save with no further model call, so it never moves when nothing moved.

Two things about it are easy to get wrong. It covers only what the CV prints, so an entry you hid is not evidence. And two scores from two adverts are not comparable, because each set its own weights. It shows where one application is thin. It predicts no employer's decision, and nothing here can promise an applicant tracking system will accept your CV.

Questions people ask

How is this different from a builder that just produces a document?
A document builder hands you a file, and from then on the file is where your career lives. Improve it and you have improved one file.
The test worth applying is what one correction buys. Rewriting buys one file. Aiming buys every application you will ever send.
If I change something on one CV, do my other CVs change?
It depends which kind. Words — a date, a job title, the phrasing of a bullet — belong to the record, so every CV containing that entry updates.
Decisions about presentation — what you hid, the order you chose, the bullets you kept — belong to that CV alone. That is what lets one record support a technical CV and a management one at once.
Do I have to fill in everything before I can make a CV?
No. The record grows as you use it and a CV shows what exists. Importing the CV you already have is the quickest start, since it fills in roles, dates and bullets at once; what that document never had room for can follow later.
Can I keep CVs in more than one language?
Yes. Each CV is set to a language and renders the content you keep in that language, translated entries sitting alongside the source rather than replacing it. A language you have written nothing in has nothing to render, so translating an entry is work done once — but the result is its own record from then on, and a later correction to the original does not travel into it.
Do I have to use the AI features?
No. Keeping a career record, running several CVs from it, choosing templates, curating what each shows and printing the result all work with nothing a model wrote.
Where it is used it does specific jobs: reading an uploaded CV into structure, reading an advert into requirements, drafting and translating. Those send CV text to third-party providers, and more than one — creating a CV from an advert alone sends text over several separate calls. The privacy notice names each provider and says what happens to the content.