Understand every point in your report.
Parse/readability 30%: contact email, damaged characters and detected layout warnings.
Structure 25%: experience, education, skills and summary headings; length prompts below 180 or above 1,200 words.
Impact 25%: equal credit for action verbs and numbers across bullet points. Numbers are evidence prompts, not proof of achievement.
Clarity 20%: deduct 8 points for each long line or vague phrase, down to zero.
JD match is separate: matched extracted terms / all extracted terms. English-only checks; no employer ATS simulation.
Illustrative example using this checker’s published rules. File warnings and wording changes are shown separately; results do not predict an employer’s decision.
Responsible for working on the backend and improving processes.
Built 3 backend reporting APIs for the finance team.
Overall = round(Parse × 0.30 + Structure × 0.25 + Impact × 0.25 + Clarity × 0.20). JD coverage is a separate percentage and never increases the overall score. The role review also stays separate. Identical text, parse metadata and job descriptions produce identical results.
Parse starts at 100: possible columns deduct 20, DOCX tables 15, header content 10, damaged text characters 25 and a missing email 15. Structure awards 25 each for recognized experience, education, skills and summary headings, then deducts 15 below 180 or above 1,200 words. These thresholds are review heuristics, not employer requirements.
Impact gives equal weight to the fraction of bullets beginning with recognized action verbs and the fraction containing numbers. Clarity starts at 100 and deducts 8 for each long line or generic-phrase match. Scores are bounded from 0 to 100. A number can be irrelevant and an unrecognized verb can still be good writing, so inspect the evidence before changing it.
Ruleset 1.1 adds evidence reviews for 12 roles without changing the general weights. The score cannot verify facts, reproduce a private ATS, assess interview performance or forecast hiring outcomes. Compare versions using the same file type; pasted text carries no original layout warnings.
Repeatable rules, visible weights and evidence behind the number.
Add a job description to see keyword matches and gaps.
Upload a text-based PDF or DOCX, or paste your text. A job description is optional.
Inspect the parsed text, score breakdown and issues tied to your original lines.
Edit a bullet, verify the facts, then re-check. Your report stays free.
English resumes only. PDF text extraction and DOCX paragraphs are checked; visual fonts, all tables, and every ATS layout behavior cannot be verified. Scanned PDFs need OCR elsewhere.
Overall = round(Parse × 0.30 + Structure × 0.25 + Impact × 0.25 + Clarity × 0.20). JD coverage is a separate percentage and never increases the overall score. The role review also stays separate. Identical text, parse metadata and job descriptions produce identical results.
Parse starts at 100: possible columns deduct 20, DOCX tables 15, header content 10, damaged text characters 25 and a missing email 15. Structure awards 25 each for recognized experience, education, skills and summary headings, then deducts 15 below 180 or above 1,200 words. These thresholds are review heuristics, not employer requirements.
Impact gives equal weight to the fraction of bullets beginning with recognized action verbs and the fraction containing numbers. Clarity starts at 100 and deducts 8 for each long line or generic-phrase match. Scores are bounded from 0 to 100. A number can be irrelevant and an unrecognized verb can still be good writing, so inspect the evidence before changing it.
Yes. You can see every score, issue and extracted line without an account. Basic checks, two-JD comparison and report printing are free.
Basic checks run in your browser. Files and job descriptions are not uploaded. AI rewriting is optional and sends only the selected text to our server and Cloudflare Workers AI after your consent.
No. Employers use different systems and rules. This is a reproducible heuristic check, not an employer score, ranking prediction or guarantee. AI is used only for optional rewriting.