Questions
Answers describe the product as it is today, in beta. Something missing? The changelog shows what has changed recently.
Importing
What do Exact, Faithful and Rebuilt mean?
They say how closely the editable copy matches your original. The match is measured after every import and shown next to the copy.
- Exact
- The same pages, every line's text identical and within 1.5 points of where it was, with no substituted fonts.
- Faithful
- The same number of pages and no missing lines, every line within 2 points.
- Rebuilt
- Anything less: the closest layout we could build, with the differences marked.
A point is 1/72 of an inch, about 0.35 mm.
Which resumes work today?
PDFs, Word files (.docx) and Google Docs links, up to 10 MB and 10 pages.
A one-column PDF printed from a browser, such as Chrome, Edge or Safari, imports Exact. In our tests, PDFs exported from Word, LaTeX and Typst, over several pages, imported Exact or Faithful on 13 October 2026, and two-column resumes with a sidebar or a divider on 14 October; other files may still come back Rebuilt. A PDF set in a font that is not open, such as Calibri or Times New Roman, is shown in a close open font, so it imports Faithful at best.
A Word file is measured against LibreOffice's rendering of it and labelled Exact, Faithful or Rebuilt, like a PDF; our sample Word resume came back Exact on 14 October 2026. Word itself may break a line differently, so look over the sheet once against your document. If LibreOffice cannot render the file, it is laid out from its source and labelled Rebuilt from Word source. A Google Docs link works when the document is shared with anyone who has the link.
A scanned resume, a PDF whose pages are pictures of text, is read with text recognition (OCR) in English and Swedish and labelled Rebuilt from a scan. A misread word can look right on the sheet, so check every line against your scan: names, numbers and dates first. If you have the original file, import that instead.
Not yet: image files, such as a photo of your resume (import the file you made it from), password-protected files (upload an unprotected copy), and old .doc, Pages, ODT or RTF files (save them as .docx or PDF).
Not supported yet: Arabic, Hebrew, Chinese, Japanese and Korean script. Such a resume imports with every line editable, labelled Rebuilt with a note that its script is not supported yet.
Not tested yet: LinkedIn's Save to PDF and resumes made in design tools. These may import Rebuilt or with lines missing; the compare view shows which.
What about two-page or two-column resumes?
Two-page resumes import, the compare view shows every page, and the check before export compares the page count with the file you imported. Two-column resumes with a sidebar or a divider imported Exact or Faithful in our tests on 14 October 2026, and text you edit wraps inside its own column. Other two-column designs are less tested; the compare view shows how close yours came.
Can I edit the copy after importing?
Yes. Click any line and type: the sheet redraws as you type, and the page count and the room left follow within a second. Scale the layout to fit the page, or zoom in to check a line. Every save becomes a version you can go back to.
Tailoring
Where do the changes come from?
From your resume and the facts you confirmed, nothing else. Each changed line lists the facts behind it. A line without a confirmed fact is flagged, and the checks before export show it before you send anything.
What is the fit rating?
Before drafting, the posting is compared with your resume and rated from 1 to 10, with reasons. At 8 or more the suggestion is to apply; at 6 or 7 you see the gaps to weigh first; below 6 the suggestion is to spend your time elsewhere.
It is advice for you. No employer sees it, and it does not predict what an employer will decide.
Where do the salary and interview figures come from?
From a web search for the employer, the role and the place in the posting, run when the posting is assessed. Each point is numbered to its source, and the sources are listed with the site and the date each was seen.
Each panel is labelled “From sources” or “Estimate”. A salary range counts as from the sources only when they state it. When nothing specific to the role is found, the panel says so and estimates from the employer type. A search is reused for seven days for the same employer, role and place.
It is guidance for you, not a promise of what you will be offered.
Is it written by AI?
Drafts are written with the help of a language model, and you review every change before it becomes a version. The model sees your text with your name, contact details, links and photo removed.
Will it get me interviews?
No tool can promise that, and we don't. RoleFramework helps you send a resume that still looks like yours, says only what is true and fits the posting. The decision is the employer's.
Languages
Does it work in Swedish and English?
Postings in Swedish and English are read, their language is detected, and the languages they ask for are flagged in the fit rating, which also suggests whether to apply with a Swedish or an English resume.
To make that resume, choose “Make a Swedish variant” or “Make an English variant” in the editor. You get a new version in the other language: headings, dates and places are translated by rule, and bullets and paragraphs by the language model, each checked for lost numbers and names. Titles, employers and skills stay as written.
The app and this site are in English and Swedish. Both open in your browser's language; to change it, use the language menu (the globe, with EN or SV) at the top of the page. Your choice is remembered in this browser, and the app also saves it to your account.
Photos and applicant tracking systems
What happens to the photo on my resume?
The import keeps the photo and its size. Your photo is never sent to the language model.
Should my resume have a photo?
It depends on where you apply. In Sweden a photo is common but optional. Many employers in the US and the UK prefer resumes without one. When unsure, leave it out.
Do the exports work with applicant tracking systems (ATS)?
They are ATS-friendly. PDF exports are text-based: every line is real, selectable text in embedded fonts, and icons are decorative. Reading order is kept column by column. The open-source PDF parsers poppler, PyMuPDF and pdfminer read them back as completely as the original, or better. A Word export keeps your file's heading styles.
In a two-column resume, the order depends on the parser, as with any PDF; the text in our exports runs one column whole, then the other, never mixed line by line.
We tested these parsers, not any applicant tracking system itself, and we give no ATS score, because each employer sets up its system differently. Before export, a check confirms that every word software extracts is visible on the page. How we tested
Your account and data
How do I log in?
Continue with Google where the log-in page offers it, or use your email address and a password. When you create an account, we email you a 6-digit code; enter it on the same page and you're logged in.
“Forgot password?” emails a code that works once, for 10 minutes; enter it, then choose a new password. Prefer no password? Choose “Email me a log-in link instead”; that link works once, for an hour.
What happens to my data?
Your resumes, postings, facts and applications are stored in your account, and the database itself refuses access from any other account. Text sent to the language model has your name, contact details, links and photo removed.
Nothing is sold or used for advertising, and this site runs no analytics. The privacy notice has the details.
How do I delete my data?
Delete any resume, with all its versions, or any application from inside the app; it is removed from the database.
To remove everything at once, open Account and data in the account menu and choose “Delete my account…”: it deletes your account with everything stored for it and logs you out. “Download my data (ZIP)” on the same page gives you a copy first. What each model call cost and a security log are kept; neither holds resume or posting text.
Where is my data stored?
RoleFramework is in a closed beta today. For the public beta, the database and files are planned in Stockholm and the web app in Frankfurt. Requests to the language model go to Anthropic in the United States.
Price
Is it free?
Free during the beta, and no payment details are asked. After the beta, RoleFramework is a paid subscription; beta users hear the price before the beta ends. Nobody is charged without choosing to subscribe.
What will the subscription include?
Everything in the beta, plus cover letters, screening answers and help with imports that come back Rebuilt, once those are built.
What is coming
What is planned next?
Planned, not built yet. What beta users ask for decides the order:
- Cover letters and screening answers.
- Help with imports that come back Rebuilt.
- How often each resume version you sent gets a reply.
Start with the resume you already have.
Free during the beta, then a paid subscription. No payment details are asked, and you hear the price before the beta ends.