Skip to content

Tailor your CV to a job description with AI

AI can tailor a CV to a job description: it reads the posting, compares it with your CV and rewrites toward it. CitedCV does this only from facts you confirm, after you buy any credit pack.1

Written 3 October 2026, updated 22 October 2026. CitedCV. How we write our guides

Start with the job posting

Paste the posting here. It is kept in this browser until you open the job match, and nothing is sent until you press analyse there.

What it means to tailor a CV to a job description

In the UK, Ireland, the EU and Turkey the document is a CV; in the US and Canada it is a résumé. This page says CV for both. To tailor a CV is to choose and order what is already true about you so that one posting's requirements are easy to see. Harvard's career office, MIT's career office (CAPD), Prospects and the UK National Careers Service all advise it for each application.

Three ways to do it

By hand (steps below). With ChatGPT, Claude or Gemini: paste the posting and your notes. The model writes from what it knows about that kind of job and can add skills and numbers you never gave it; MIT CAPD warns of made-up experience, which researchers call hallucination (Huang and colleagues, arXiv, 2023). With a builder that reads the posting and your facts, as CitedCV does: tools differ in what the writer may add.

What reads a tailored CV

Applicant tracking systems (ATS) such as Workday, Greenhouse, Lever, Taleo, SuccessFactors and iCIMS parse, store and search a CV. Greenhouse's Talent Matching compares resume terms with skills a recruiter chose; its support page says it does not auto-reject or auto-advance anyone. A person decides, and none detects AI text.

What you can do in the first minute, and what it costs

The quickest useful step is a job match: one posting against one CV, shown line by line. It costs 1 credit and needs a purchased credit pack. Reading the posting first costs no credits.

StepNeedsCost
Read the posting from a link, screenshot, file or pasted textAny credit packNo credits; you see and trim the text first
Job match, with the evidenceAny credit pack1 credit; the same posting and CV are never charged twice
Quick rewrite for the postingAny credit pack6 credits, no questions
Full rewrite for the postingAny credit pack10 credits, asks about gaps
Compare again after a rewriteAny credit packNo credits the first time after a rewrite from the report, otherwise 1 credit
Export PDF or DOCXAny accountNo credits; a premium template costs 5 credits once

Packs, each paid once: Starter $9 for 30 credits, Pro $29 for 150. A pack never renews and its credits do not expire (pricing page). A match plus a full rewrite is 11 credits, so Starter covers two.

On a free account

A new account gets 6 credits and needs no card: enough for one quick draft of a general CV. Auditing a CV you already have and fixing its lines cost nothing. Reading a posting, matching and a CV written for a posting open once you buy any credit pack; the free signup credits do not unlock them.

By hand, at no cost

  1. List the posting's required and preferred requirements.
  2. Beside each, write the CV line that shows it, or "nothing".
  3. For each "nothing", add a line only if you did the work.
  4. Move the closest-matching experience to the top and cut the rest.

How tailoring works in CitedCV

  1. Give it the posting as a link, screenshot, file or text, and check what was read. LinkedIn, Indeed and Glassdoor links are not read (those sites need a login); paste the text.
  2. Pick the CV: one made in CitedCV, a file up to 5 MB, or pasted text. The report takes about half a minute.
  3. Read the report. Each requirement is shown with the CV line that covers it, or with "No line on this CV shows it".
  4. Rewrite for the posting. A CV made in CitedCV can be rewritten at once; an uploaded one is first imported (reading costs no credits) and each line confirmed.
  5. Compare again to see the score before and after.

What a match report contains

  • A score from 0 to 100 and a band: fits, close or far.
  • Decisive requirements met (such as 6 of 8), plus requirements covered and search terms present as percentages.
  • Each requirement as Required, Preferred or Assumed, with the verdict Shown, Partly or Not on the CV and the CV line, decisive words highlighted.
  • A flag where a term is on the CV but no line shows the work, and missing search terms ("Add one only if it is true").

What a rewrite does about a gap

A rewrite uses only facts you confirmed, and brings onto the CV a confirmed fact that covers a missing requirement. If nothing on file covers it, a full rewrite asks you first (up to three questions by default); a declined answer stays out, and a quick rewrite asks nothing. The step that writes the bullets never sees the posting, only your facts, and a verifier checks every number, date and name. A line that still fails after a rewrite is dropped.

Proof: how a report is computed and what we measured

Every figure comes from CitedCV's code or our own runs, and says which.

WhatFigureSource
WeightsRequirement kind: required 1, preferred 0.4, assumed 0.15, times the requirement's own 0 to 1 weight. Verdict: shown 1, partly 0.5, term only 0.3, nothing 0Scoring code, match-score@1
Score and bands85% requirements, 15% search terms; 75 or more fits, 50 to 74 close, under 50 far; a missing decisive requirement caps it at 74Same
EvidenceA quote counts only if it is found on the CV line it names, ignoring case and spacingSame, checked in code
One run of ours21 CV lines, 15 requirements, 12 of 12 quotes verbatim on the CV, 31 secondsOur staging run, 2026-09-18; one run, not a benchmark

Check a report yourself

Read the evidence line for three Shown requirements and ask whether you could defend each in an interview.

Who this is not for

  • You apply to many similar roles and want a tailored file in seconds. RecastCV and Enhancv describe tailoring in under 30 seconds by their own figures (2026-09-08); a full CitedCV rewrite takes minutes.
  • You do not want to pay anything. Use the by-hand steps, or Teal, Rezi or ResumeFab, which have free tiers, and check every line.
  • You want the posting's terms inserted whether or not you did the work. Rezi lists missing keywords and can generate a bullet for each one you accept; accept only what you can back up.

Alternatives, and whom each fits

CitedCV is ours, so read this as a disclosure. Competitor rows repeat each tool's own pages as of 2026-09-08.

OptionWith a postingFits
By handYou read the requirements and edit your own CV.The one or two applications that matter most
ChatGPT, Claude, GeminiYou paste posting and notes; the model predicts the likeliest CV for that kind of job.A first draft you check line by line
RecastCVRewrites a master CV in the posting's language, shows an ATS match score, returns a .docx; one credit per tailored CV.A CV you trust, many similar postings
EnhancvOne-click tailoring, each change highlighted to accept or reject; says it uses only your real experience.A resume you have, many layouts
TealFree job-search workspace; scores a resume against a posting and rewrites toward it; no stated rule against numbers you never supplied.Tracking many applications

Limits we state out loud

No CV tool can promise an interview, and none controls what Workday, Greenhouse, Lever, Taleo, SuccessFactors or iCIMS do after you apply. They do not detect AI text. The score describes this posting and this CV, not any employer's software.

A language model judges each requirement. The quote is verified in code, but whether it shows the requirement is a judgement, so read the line. If that step fails you get a flagged term-matching report and no charge. A page built only in the browser can come back empty; paste the text. Reads are capped at 40 links and 30 pictures a day, and 20,000 characters of posting.

With thin facts the CV is shorter, not padded. CitedCV does not submit applications for you.

Questions people ask

Is there an AI resume builder based on a job description?

Yes. CitedCV, RecastCV, Teal, Rezi, Enhancv and ResumeFab take one, and ChatGPT, Claude and Gemini do if you paste it. They differ in what the writer may add: a chat model can add skills you never gave it; CitedCV uses only facts you confirmed, after you buy any credit pack.

Does tailoring mean copying keywords from the job description?

No. Use the posting's term once, where you did the work; a term you cannot discuss in an interview is a risk. CitedCV never adds a skill from the posting that your facts lack.

How much does it cost to tailor a CV in CitedCV?

From 1 credit: a match costs 1, a quick rewrite 6 and a full rewrite 10, and reading the posting costs none. All need a purchased credit pack, from $9 for 30 credits, paid once.

Can I tailor a CV on a free CitedCV account?

No. A free account has 6 credits, enough for one quick draft of a general CV. Matching and a CV written for a posting open once you buy any credit pack.

Can I give CitedCV a LinkedIn job link or a screenshot?

A screenshot, yes; a LinkedIn link, no. LinkedIn, Indeed and Glassdoor links are not read because they need a login. Paste the text or upload a screenshot, and check it for a wrong letter.

Will a tailored CV get me an interview or past an ATS?

No tool can promise either. Workday, Greenhouse, Lever, Taleo, SuccessFactors and iCIMS parse and search a CV, none detects AI text, and a person decides.

Sources

  1. Harvard Office of Career Services, AI for resumes and cover letterscareerservices.fas.harvard.edu/ai-resumes-and-cover-letters/
  2. MIT Career Advising and Professional Development, Career toolkit: crafting an effective resumecapd.mit.edu/resources/career-toolkit-crafting-an-effective-resume/
  3. MIT Career Advising and Professional Development, AI uses for resume writingcapd.mit.edu/resources/ai-uses-for-resume-writing/
  4. Prospects, How to write a CVwww.prospects.ac.uk/careers-advice/cvs-and-cover-letters/how-to-write-a-cv
  5. National Careers Service (GOV.UK), How to write a CVnationalcareers.service.gov.uk/careers-advice/cv-sections
  6. Greenhouse Support, Talent Matching FAQsupport.greenhouse.io/hc/en-us/articles/41131886674075-Talent-Matching-FAQ
  7. Huang et al., A survey on hallucination in large language models (arXiv 2311.05232, 2023)arxiv.org/abs/2311.05232