How our resumes parse,tested in the open.

1. 01

   Title and company on separate lines

   The role heading was one line, “Senior DevSecOps Engineer — Northwind Cloud Services”, which OpenResume read as a company with no title. The company now sits on its own line under the title, in every template and in both PDF and DOCX.

2. 02

   Signal’s section numbers are drawn, not typed

   OpenResume looks for a heading that is one bold, all-caps word, and Signal’s “01 SUMMARY” wasn’t one. In the PDF the numbers are now vector outlines in the same font and color, so the text is just “SUMMARY”. Word has no ATS-safe way to draw them, so the DOCX leaves them out and keeps the heading at the same indent.

3. 03

   Name first, tagline under it

   The small focus-area line (“PLATFORM ENGINEERING · KUBERNETES · DEVSECOPS”) sat above the name, so parsers that take the first line as the name got the tagline. It now sits under the name. In Signal it also moved a few pixels lower, because pdf.js joined it to the 32pt name when it was closer.

4. 04

   Education on two lines

   “B.S. Computer Engineering — University of North Texas, 2016” is now the degree on one line and the school on the next, with the year right-aligned like role dates. The words are unchanged; the line is only split.

Fonts, colors and spacing are otherwise unchanged. Separate lines make roles a little taller, so some templates now run onto a second page.

• These are not the systems employers use. Workday, Greenhouse, Lever, iCIMS, Taleo and the rest run their own proprietary parsers. We can't test those for free, and we won't guess how they behave. Treat these results as a sign of how a file holds up in general, not as a prediction for any one employer.
• Open-source parsers are simpler. OpenResume states it only handles single-column English resumes; pyresparser was last released in 2019 and relies on a small spaCy model. Commercial parsers are usually more forgiving, but not always.
• One resume. A different person's history (more roles, a second degree, no dates) could parse better or worse.
• Parsing is not ranking. Even a perfect parse says nothing about how a recruiter or a search filter scores the content. That is what keyword coverage is for.