Resumes in any layout, normalized in one shape.

Resume parsing — also called CV parsing or résumé extraction — is the process of turning a free-form resume document (PDF, DOCX, image) into structured candidate data. Every applicant tracking system (ATS) and sourcing platform sits on top of this layer. Done with regex and templates, parsing breaks the moment a candidate uses a non-standard layout; done with a layout-aware engine, the same schema covers thousands of variations without per-template setup.

On the sample resume above, Ztract returns the candidate's full name, email, phone, and location; the headline summary paragraph; the work experience array with each role's title, company, and start/end dates; the education array with school, degree, and major; and the skills array — each skill as its own clickable chip linked back to where it appears on the page. The sample resume is for a UX Designer with three roles and two degrees; the output is ready to ingest into Greenhouse, Lever, or Workday in seconds.

Two design choices matter here. First, skills come back as a typed array of strings — not a comma-separated blob — so downstream skill-matching can compare against a candidate database without further string-splitting. Second, dates are normalized: 'Jun 2023', '06/2023', '2023-06', and 'Present' all land in a sortable format. Multi-column resumes, sidebars, and 'creative' layouts that confuse most parsers are read in the right column order.