AI résumé tailoring you can audit line by line

How a review runs

1. Pick a live role

   From the catalogue of 285 companies' own boards, so you are tailoring against a posting the employer currently has open — not a description pasted from somewhere else. The server reads the description from the catalogue, not from your browser.

2. Run Analyze & tailor

   You get a fit assessment with reasons and gaps, suggested résumé edits, and a cover-letter draft. Missing qualifications go in the gaps list, not into the résumé.

3. Approve edit by edit

   Each suggestion is shown against the exact sentence it replaces. Reject one and the rest stand. Your original file is untouched throughout.

The four checks that run before you see anything

• Every edit must quote your résuméA suggested edit has to quote an exact, unique passage of the file you uploaded. If the quoted original is not found verbatim, or is found twice, the edit is discarded before you ever see it — which removes the whole class of edits that invent a new bullet point.
• Invented figures are strippedA replacement may not introduce a number that was not already in the passage it replaces. Percentages, headcounts, latency and cost figures that appeared from nowhere are dropped mechanically, not by asking the model nicely.
• No overlapping rewritesIf two suggestions touch the same sentence, applying both would produce a sentence neither proposed. The overlap is dropped, so an accepted edit means what it said it meant.
• Your original is never rewrittenThe file you uploaded stays as you uploaded it. You approve edits one at a time, and changing your profile details invalidates an earlier approval so a stale review cannot follow new facts.

About match scores

The match number beside a role, and the fit assessment an AI review returns, are SchoolWhool's own reading of the posting against your profile. Neither is an employer ATS score, and neither predicts whether you will be hired.

Use a score the way you use a spellchecker's underline — as a prompt to look again, not as a target. Past a point, raising a relevance number means stuffing vocabulary into sentences a human has to read, and the human is the one who decides.

Test any tailoring tool in ten minutes

Including this one. Run it on a free tier before you subscribe to anything.

• Use your real résuméA sample résumé will not expose invention, because nothing in it is at stake.
• Pick a posting with a genuine gapIt should ask for at least one thing you have never done.
• Read every change against the originalNot the summary of changes — the changes. Look for a skill that appeared, a number that appeared, 'contributed to' upgraded to 'led', and a team project that became yours.
• Find your original the next dayIf the only copy is the tailored one, the tool has overwritten your source of truth and every future session compounds on a rewrite.