Why a PDF resume fails an ATS parse
Most “ATS rejected my CV” stories are a parse failure. The file did not yield a name, dates, or a readable experience section, so a recruiter search never finds you.
An applicant tracking system stores the text it can extract. It does not score you out of 100. If select-all in the PDF pastes nothing, or pastes your jobs in the wrong order, that is the failure. The checks below are the ones that show up on DevOps and cloud CVs that were designed as posters.
The text is a picture
A CV exported from a design tool as a flattened image, or a scan of a paper page, has glyphs a human can see and a parser cannot read. Open the PDF, Select All, copy, paste into a plain text editor. If the paste is empty or a few broken characters, rebuild from a text template and export again.
Contact details live in the header or footer
Some exporters write the name, email, and phone only in the page header. Parsers often skip headers and footers, so the body looks like an anonymous work history. Put name and contact in the normal flow of the first page, then repeat them in a header if you want them on page two.
Two columns are read as one scrambled line
A sidebar of skills beside a main column of jobs is fine for a person. A parser walks coordinates. It can interleave “Kubernetes” with the middle of a bullet, or read the sidebar before any employer. One column, standard headings (Experience, Education, Skills), and dates on the same line as the role survive that walk.
Icons and bars replace words
A skill bar, a flag icon, or a logo with no adjacent text does not become the word “AWS” or “German”. Write the language, the tool, and the level as text. Decorative icons can stay; they must not be the only place the fact exists.
The longer explanation of what an ATS stores, and what it does not: What an ATS actually does to your CV
DOCX versus PDF, and when each is the safer export: Resume file formats that parse