See what an ATS actually reads on your CV

What does the ATS actually see?

An applicant tracking system does not look at your CV. It reads it. Before any recruiter opens your file, the ATS runs your PDF through a text parser that strips away fonts, colours, columns, and spacing, and keeps only a stream of raw text. That stream is what gets indexed, searched, and matched against the job description. If your name, your job titles, or your skills are missing or scrambled in that stream, they are invisible to every search a recruiter runs.

That is exactly what this checker shows you. When you upload your CV, we extract the text the same way parsing software does and display it back to you as the ATS View: the literal, unstyled text a machine reads from your file. No simulation, no mockup. If a word is shattered into fragments because of aggressive letter spacing, or a whole sidebar is missing because it was exported as an image, you will see it with your own eyes.

Most CV checkers give you a score and ask you to trust it. We give you the evidence first and the score second. The ATS View is the proof; the checks below explain what we look for in it.

How is the report built?

The report is 100 per cent deterministic. Your PDF is parsed on our server with the same open-source extraction approach used by parsing pipelines, then a fixed set of rules runs against the extracted text. No AI model reads your CV, so the same file always gets the same result, and every finding can be traced to something visible in the ATS View. The report runs four named checks:

• Text extraction. Did your PDF yield clean, machine-readable text? We measure the share of well-formed words in the output. Files that produce shattered fragments (a common side effect of decorative letter spacing or text baked into images) fail here, because a parser that cannot read words cannot match keywords.
• Contact information. Can a parser locate an email address and a phone number in the extracted text? These are the two fields recruiters use to reach you, and headers rendered as graphics are a classic way to lose them.
• Section headings. Do standard headings like Experience, Education, and Skills appear in the text stream? Parsers use these labels to segment your CV into fields, so unconventional or missing headings degrade how your data is filed.
• Dates. Are employment dates present and recognisable? Date ranges let systems reconstruct your work history and compute tenure; a CV without parseable dates looks like a CV without a timeline.

How is the score calculated?

The score is a weighted sum of the four checks, and we publish the weights instead of hiding them: Text extraction counts for 40 points, Contact information for 20, Section headings for 20, and Dates for 20. A check that raises a warning rather than a failure earns half its weight. There is no secret ingredient and no randomness: fix a finding, rescan, and the score moves by exactly the documented amount.

One honest caveat: no external tool can know which ATS a specific employer runs or how its recruiters configured it. What a checker can verify is whether your file survives parsing intact, and that is the part that is common to every system. That is what we score, and nothing else.

What are Opportunities, and why don't they affect the score?

Below the checks, the report lists Opportunities: writing suggestions that make your CV stronger for the human who reads it after the machine. They cover cliches to cut, bullets that would benefit from numbers, weak verbs to replace, a summary that runs past 60 words, inconsistent date formats, missing contact fields, and overall length outside the 350 to 900 word range that fits a one to two page CV.

Opportunities never change your score, by design. Mixing parsing problems (machine facts) with writing style (editorial judgement) into a single number is how other checkers manufacture urgency. We keep them separate so you always know whether an issue would actually block a parser or simply reads poorly.

If you paste a job description alongside your CV, the report also runs a plain term-frequency match against the advert, so you can see which of its recurring terms your CV already covers. Again: transparent counting, not AI scoring.

Is it really unlimited?

Yes. There is no account, no email gate and no scan counter on the report. You can scan a CV, edit it, and rescan as many times as you need until every check passes. Your file is parsed in memory and never stored; when the response is sent, the document is gone.

Why keep it open to everyone? Because the checker is how we prove the rest of the product. Every one of the 25 ATS-Verified templates in our builder passed this same style of extraction testing before earning its badge, and we published the full numbers in our 2026 template study. If the checker convinces you your current CV needs rebuilding, the builder is there; if your CV already passes, you leave with that confirmation in hand, and we are fine with that.

What should you do with your results?

A practical order of operations after your first scan:

1. Read the ATS View before the score. Confirm your name, titles, employers, and skills all appear as intact words.
2. Fix any failed check first: those are parsing problems, and no amount of good writing compensates for text a machine cannot read.
3. Work through the Opportunities next, starting with quantification: bullets with numbers survive skim-reading by humans.
4. If you are targeting a specific role, paste the job description and close the gaps on terms you genuinely have. See our guide on tailoring your CV to a job description.
5. Rescan after each round of edits. The checker is deterministic, so progress is real, not noise.