Greenhouse Resume Parsing Mechanics
Greenhouse is unusually open about what breaks its resume parsing. Its support center lists the causes of a failed parse, and explains how its AI matching uses your resume. This guide sticks to what Greenhouse has published, with links.
1. What parsing does
When a resume is added, Greenhouse Recruiting scans it and auto-fills the candidate fields it can detect. When a parse fails, those fields stay empty, and the recruiter has to work from your file instead.
2. What Greenhouse says breaks a parse
Greenhouse's own list of causes:
- File size over 2.5 MB.
- A resume uploaded as an image rather than a document. Use a .docx or a PDF with real text.
- Formatting: spaces between letters, graphics, photos, word art, complex tables, headers and footers, and columned layouts.
- Inconsistent sections. Use the same structure throughout, with standard headings.
- Abbreviations. Abbreviated job titles (“Sr. Account Exec”) and company names without “Inc.”, “Co.”, “LTD” or “LLC” can also cause problems. Write titles out in full.
Almost all of these are covered by the ATS resume format guide, and the parse test catches several of them in your own resume.
3. AI Talent Matching
Greenhouse's Talent Matching feature uses details parsed from your resume: skills, years of experience, job titles, employment dates and company names. It sends them to a third-party AI service (OpenAI) for processing. Greenhouse says it uses an “embedding representation of skills and job titles” to run a semantic search, so related terms can match without being identical. Recruiters see a match category (Strong, Good, Partial or Limited).
It doesn't decide for the recruiter. In Greenhouse's words, “there is no mechanism that automatically takes action to advance or reject a candidate based on an AI rating,” and humans make the decision on every candidate.
What helps: clean parsing of your titles, dates and employers, and skills stated plainly in your experience. Matching is only as good as the fields it's given.
4. Fraud detection
Greenhouse's fraud detection looks at signals around an application, such as an IP address linked to a data center rather than a home connection, which it calls “strongly associated with fraud.” It sends a fraud report to the employer rather than rejecting anyone automatically.
What isn't public
Greenhouse doesn't name the parser behind its built-in parsing, and doesn't describe how it splits a resume into sections. It does support third-party parsing integrations, so a given employer's setup may differ.
Frequently Asked Questions
Why might Greenhouse put my experience under the wrong section?
Greenhouse lists inconsistent sections among the causes of a failed parse. Use the same structure throughout and standard headings such as Experience, Education and Skills; a creative label is easy for a person to read and easy for a parser to miss.
Can icons in my resume header break parsing in Greenhouse?
They can. Greenhouse lists graphics, complex headers and footers, and spaced-out letters among the formatting that breaks a parse. Icon fonts also turn into unknown characters when text is extracted. Write contact details as plain text in the body of the page.
Does Greenhouse's AI reject candidates automatically?
No. Greenhouse says there is no mechanism that automatically advances or rejects a candidate based on an AI rating. Talent Matching gives recruiters a match category; a person makes the decision.
Sources
Check your resume before you apply on Greenhouse
Tailor a single-column resume with full job titles and standard sections from your real experience in the Axiom Workstation.