How we score, and why

We never invent your experience

Every suggestion has to be grounded in something already on your resume. When it can't be, we withhold it — because the cost of a fabricated line lands on you.

This is the rule we're least willing to bend: AI review will not invent experience for you. Every suggested edit has to be grounded in something you already wrote. When a suggestion can't be, it doesn't reach you.

It's enforced, not just requested. We don't rely on politely asking a model to behave — suggestions are checked against your own content before you ever see them, and ungrounded ones don't pass.

What this means in practice

  • A rewrite reshapes what you wrote. It doesn't add a fact.
  • If a posting wants something you haven't done, you'll get no suggestion for it — not an invented one.
  • If we can't tie a proposed improvement back to your resume, you'll be told some suggestions were held back rather than handed a plausible-sounding claim.
  • Your cover letter is drafted the same way, from facts already in your resume.

Why we're this strict

Because the harm lands on you, after you've accepted the job, when it's hardest to recover from.

SHRM's reporting on resume dishonesty found that of people who lied and were hired, 41% had the offer rescinded once it was discovered and 18% were terminated later. Nearly one in three Americans admit to lying on a resume at some point.

And it gets caught. HireRight's 2025 benchmark of over 1,000 HR and screening professionals worldwide found more than 75% of employers caught at least one candidate discrepancy during background screening; nearly two in five find one in every twenty candidates. Employment verification — your job titles, dates and employers — is the single check most likely to turn something up.

A tool that quietly padded your resume would be handing you a liability with a delayed fuse.

The environment is getting less forgiving, not more

Employers know AI has made fabrication easy, and they're tightening up.

Greenhouse's 2025 research found 28% of job seekers admit using AI to generate fake work samples and 32% have claimed AI skills they don't have. On the other side, 74% of hiring managers are more concerned about fake credentials and misrepresented experience than a year ago, 65% have caught applicants using AI deceptively, and 39% of US hiring managers are running more in-person interviews specifically to verify that people are who they say they are.

Gartner forecasts that by 2028 roughly one in four candidate profiles worldwide will be fake, and recommends employers communicate their fraud-detection efforts and the consequences of fraud.

Verification pressure is going up. Anything invented on your resume today is more likely to be checked tomorrow than it was last year.

The cost isn't only the offer

Among people who embellished and were hired, reported consequences include ongoing stress about being found out, the overstatement becoming obvious on the job, and struggling because the skills didn't match the resume. That comes from a background-check vendor's survey — treat the exact figures cautiously — but the shape of it is intuitive and matches what the SHRM data shows.

You'd also be walking into interviews defending a line you didn't write. That's a bad position on its own.

Withheld, not replaced

When a suggestion is withheld you're told, in a quieter tone, that some were held back. It's never dressed up as a completed round.

Claiming we'd finished when we'd withheld work would be exactly the confident-but-untrue statement this whole guarantee exists to prevent. It would be odd to enforce honesty in your resume and abandon it in our own interface.

Detail: why a fresh pass can come back empty.

What we won't publish

We won't describe how the grounding check makes its decisions — what it looks at, what triggers it, or how it compares a suggestion to your text.

That's the one piece of reticence in our approach we'd defend hardest. An article explaining exactly how the guard decides is an article explaining how to slip something past it, and the person harmed by that would be the user who did.

The honest limit

We can stop the tool inventing things. We can't stop you — you can type anything you like into your own resume, and we won't know.

What we can promise is that nothing we hand you will be a claim you didn't earn. Everything AI review proposes is your own experience, said better.

Related: writing bullets that land and reading your keyword chips.

Still need help?

If this didn’t answer your question, you can contact us. We’ll reply by email.