An honest AI resume builder — no fake experience

Why hallucinated bullets get you fired

The first time you see an AI-tailored CV, it's startling how good it looks. Skills match the JD, numbers sound impressive, and seniority feels right. Then you read carefully and notice: half of those numbers aren't in your original CV.

Adding a skill you don't have gets you past the ATS. It doesn't get you past the interview, where the hiring manager will ask for specifics. It also doesn't get you past a background check at any sizeable company — employment verifiers check dates, titles, and credentials against your stated employers, and a fabrication that surfaces six weeks after start date is a fast path to termination.

And the dishonest version doesn't even help you long-term: even if you somehow land the job, you'll spend the first three months hiding skills you claimed. That's a brutal way to start.

The honesty check, line by line

AdaptMyCV's pipeline produces three sections in parallel — header summary, skills, experience bullets — and then runs a separate AI pass whose only job is to score the output for honesty. That pass receives the source CV and the rewrite side by side. For every line in the rewrite it asks: is this grounded in the source?

The check reports four categories:

• Grounded — the claim is directly supported by a specific line in your source CV.
• Defensibly inferred — the claim isn't a direct quote but a hiring manager would assume it. Example: you wrote “led Agile sprints” → “Scrum” lands in the skills list. Defensible.
• Unverified — added by the model but not supported. Shown to you with a warning. Default behaviour: keep it but flag it so you can edit before download.
• Blocked fabrications — invented companies, changed dates, fake degrees, fabricated numbers. Removed before you see the output, with a note explaining what was caught.

On top of that, a deterministic check enforces a hard rule the model occasionally breaks: never downgrade a qualification to match a JD's minimum. If your CV says English C1 and the JD asks for B2, the output stays at C1 (sometimes phrased “C1 — exceeds B2 requirement”). See the about page for the full implementation detail.

What we won't do

• Won't invent companies or roles. Your work history is verbatim. No new employers, no shifted titles.
• Won't change dates. Date ranges in your experience section are immutable.
• Won't fabricate numbers. Percentages, revenue figures, team sizes, growth rates — only what you already wrote.
• Won't add degrees or certifications. No new AWS, Azure, GCP, PMP, Scrum Master, or any other credential gets appended.
• Won't downgrade levels. Language proficiency and seniority go up or stay the same — never reduced to match a JD floor.

What we will do

Tailoring is genuinely useful when it's honest. We:

• Reorder your existing skills so the JD's priority terms appear first.
• Rewrite bullets to use the JD's language for the work you already described (“dashboards” → “A/B testing dashboards” if you mentioned A/B testing).
• Surface adjacent terminology that a hiring manager would assume from your role (Agile → Scrum / Kanban / retrospectives).
• Compress and tighten language so each bullet leads with the outcome, not the tool.
• Run an ATS-keyword coverage report so you can see which JD terms made it onto the page.

Pricing — pay per CV, no subscription

10 coins (≈ $2) per Adapt run. 5 coins = $1. New accounts get 10 free coins on signup, so the first tailoring is free. There's no monthly fee and no auto-renewal — you top up once and spend the balance when you actually need a CV. Unused balance is refundable within 24 hours of top-up.

See full pricing for top-up sizes (small, medium, large bundles).

FAQ