Blog
-
What a Parser Does With Your Links
Hyperlinked text, mailto links and QR codes often lose the address during extraction. How to write a link that survives being turned into plain text.
-
What a Page Break Does to a Parsed Resume
A second page adds running headers, page numbers and split job entries to the text stream. What extraction does with each, and how to keep them harmless.
-
The Characters That Extract Wrong
Ligatures, subsetted fonts, icon glyphs and smart punctuation can corrupt individual characters during extraction. How to spot and fix each one.
-
Where One Job Entry Ends and the Next Begins
Parsers have to cut your work history into separate records. What tells them where the cut goes, and what happens when the cut lands wrong.
-
When Your Resume Is a Picture
Image-based PDFs and outlined text extract as nothing at all. How to tell whether your file contains real text, and what OCR does and does not fix.
-
Hidden Instructions Aimed at an AI Screener
Burying a command in white text for a language model to obey is prompt injection on your own job application. Why it fails, and what it costs.
-
LLM Screening Versus Keyword Parsing
A language model reads your resume as text, not as fields. What that changes about formatting, wording, and evidence — and what it does not change.
-
How to Test Your Resume Like a Parser Would
Four free checks that show you what software extracts from your resume, so you can fix formatting problems before you submit anything.
-
File Formats: PDF, DOCX, and Plain Text
What each resume file format actually stores, how that affects text extraction, and which one to send when the application form doesn't say.
-
Job Titles, Dates, and the Fields Parsers Look For
How software identifies your employers, titles, and date ranges — and the small formatting choices that make a job land in the wrong place.
-
What a Resume Parser Actually Sees
A step-by-step walk through how software turns your resume file into structured fields — extraction, reading order, sectioning, and field matching.
-
Why Your Two-Column Resume Breaks
Columns, tables, and text boxes look tidy and parse badly. Here's exactly what happens to the text inside them — and what to use instead.
-
How Systems Know You Applied Before
Candidate records are deduplicated by email and name. What that means when you reapply, update your resume, or use a second address.
-
When the Application Form Autofills From Your Resume
Autofill shows you the parser's output for free. How to read those prefilled boxes as a diagnostic, and what to fix in the file afterwards.
-
Parser Confidence and the Unclassified Block
Parsers guess, and record how sure they are. What happens to text a parser cannot classify, and how to stop producing any.
-
Why the Same Resume Parses Differently
One file, two employers, two different records. The sources of variation between parsers, and how to build a resume that survives all of them.
-
The Screening Questions That Come Before Parsing
Some application decisions never touch your resume. How structured form questions work as knockouts, and why they deserve more care than your file.
-
What a Resume Match Score Measures
Match percentages are usually a coverage ratio between two term lists. How that number is built, what it cannot see, and why chasing it backfires.
-
How Recruiters Search the Parsed Record
Your resume becomes a searchable database row. What a boolean query over that row matches, and why the search happens weeks after you applied.
-
Location, Remote, and the Fields That Filter
How a parser decides where you live, why an omitted city creates an empty field, and why remote eligibility is answered on the form, not the resume.
-
How a Degree Requirement Gets Checked
Education parsing turns your schooling into fields: institution, credential, field of study, dates. How a degree filter reads them, and what breaks.
-
How Your Job Title Gets Normalised
After a parser finds your job title it often maps it to a standard title in a taxonomy. What that mapping does to unusual titles, and how to help it.
-
How Software Counts Your Years of Experience
Years of experience is usually calculated from your parsed date ranges, not read from a sentence. What that arithmetic gets wrong, and how to fix it.
-
Keyword Density Is Not What Parsers Measure
Repeating a term does not raise a score in a presence-based match. Where the density myth came from and what to do with the space instead.
-
Where a Skill Has to Appear to Be Found
A skill in a graphic, a footer, or a shredded sidebar does not exist in the parsed record. Where text has to sit to survive extraction.
-
Acronyms, Synonyms, and Spelling Variants
Why a parser may not know that K8s means Kubernetes, and the honest way to cover both forms of a term without stuffing your resume.
-
How Skills Extraction Works
How software finds skills on your resume: matching text against a vocabulary rather than understanding it, and what that mechanically rewards.