Rewriting your CV with AI without sounding like AI

What it is good at

• Quickly rewords bullets around the exact language of a job advert.
• Good at spotting missing achievements when you describe your day-to-day work.
• Can generate several tone versions, formal, plain, senior, in minutes.
• Useful second pair of eyes for structure and length before a human reads it.

Where it falls short

• Will invent plausible-sounding achievements if you let it, always verify every number yourself.
• Default output leans on the same handful of words every applicant now uses.
• Cannot judge whether you're actually a good fit; it only judges the wording.

How to use it for creative work

1. 01

   Paste the job advert and your raw history together

   Give the model the full job advert and a messy, unedited list of what you actually did in each role, including numbers. Ask it to highlight which of your experiences map to which requirement in the advert before writing anything.

2. 02

   Ban the generic words up front

   Tell it explicitly: no 'passionate', 'dynamic', 'results-driven', 'team player'. Ask for one concrete detail, a number, a named tool, a named outcome, in every bullet instead.

3. 03

   Ask for three lengths

   Request a one-page version, a two-page version and a LinkedIn-summary version from the same source material. Comparing them side by side shows you which details actually matter.

4. 04

   Write the cover letter from the CV, not from scratch

   Once the CV is right, ask for a cover letter that references only what's already in the CV, plus one sentence on why this company specifically. This stops it from inventing motivation that sounds hollow.

5. 05

   Read it aloud before sending

   If a sentence doesn't sound like something you'd say in an interview, cut it. This single check catches most of the remaining AI tone.

Ready-to-use prompts

Here is a job advert: [paste]. Here is my unedited work history: [paste]. Before writing anything, tell me which of my experiences map to which requirement, and which requirements I have no strong evidence for.

Rewrite these bullets for the role of [job title]. Do not use: passionate, dynamic, results-driven, proactive, team player, detail-oriented. Every bullet must contain one number or one named tool/outcome. Bullets: [paste]

From this material, produce: a one-page CV, a two-page CV, and a 3-sentence LinkedIn headline and summary. Keep facts identical across all three; only cut and compress.

Write a cover letter for [role] at [company] using only facts present in this CV: [paste]. Add one sentence on why this specific company, based on: [one real detail about them]. No filler opening sentence.

List every number, percentage and named outcome in this CV draft. For each, ask me to confirm it's accurate before we finalise. Flag anything that reads as invented or vague.

Questions people ask

Will recruiters notice an AI-written CV?

They notice generic ones, not AI-assisted ones. The tell is vague adjectives and no numbers, both of which are fixable regardless of what wrote the first draft.

Should I let it write my achievements for me?

No, give it your real history and let it improve the wording. If you let it fill gaps, it will guess, and interviewers ask follow-up questions you won't be able to answer.

Does this work for creative portfolios too?

The CV text, yes. The portfolio itself needs a visual review, see the portfolio review workflow for that.