Best Resume Format for AI Screening: What Parsers Need

AI screening tools do not read your resume the way you see it. They read the text a parser extracted from your file. Workday's HiredScore, for example, now uses a language model that "evaluates each screenable qualification individually against the candidate's parsed resume," according to Workday's August 2026 release note. If the parse is wrong, the AI grades the wrong resume. So the best format for AI screening is the one that parses cleanly. This page lists the specifics, taken from the vendors' own documentation.

#Why format decides what the AI sees

Every AI screener in large-employer hiring sits downstream of a parser:

• Workday's resume parsing documentation describes extracting candidate information from an attached resume into profile fields.
• Phenom says its parser extracts "years of experience, hard skills, soft skills, geographic location, and other pertinent details" (Phenom).
• Eightfold's matching model takes "titles, work experiences, education, skills, and natural language of resumes" as inputs (Eightfold OFCCP disclosure).

A layout that confuses the parser does not just look odd. It moves your skills into the wrong field, splits a job title from its dates, or loses your contact details. The AI then scores that damaged version.

#What the vendors say breaks parsing

Greenhouse publishes the most specific list. Its support article on unsuccessful resume parses says it "can't parse resumes larger than 2.5MB." It names these causes:

• resumes uploaded as images
• "resumes with spaces between the letters"
• graphics, photos or word art
• tables, headers or footers
• contact information in headers, footers or text boxes
• columned layouts without clear sections
• abbreviated job titles.

Workday's parsing page adds: "For best results, use resumes that don't have images or image-based styles." It also notes that Workday "doesn't auto-fill fields you configure as hidden or these fields: Languages, Skills." So on some Workday forms you may need to enter skills by hand even when they are in your resume.

Phenom's own blog warns that resumes "in a format not easily parsed by NLP (e.g., image-heavy resumes or files scanned in as images)" are harder to match (Phenom).

#Reverse-chronological still wins

AI matchers compute experience from dated roles, and some weigh recency. Eightfold's engineering blog describes distinguishing recent from historical skill use and modelling title progression. A functional resume, which groups skills without tying them to dated jobs, gives those signals nothing to work with. List roles newest first, each with a title, employer, location and month-year dates.

Keep section names standard. "Experience," "Education" and "Skills" are what parsers are built to find. "Where I've Made an Impact" is not. Our guide to ATS resume section headings lists safe alternatives.

#The skeleton that parses

Here is the full structure in plain text. If your resume, saved as plain text, looks like this, the format is doing its job:

FIRST LAST
City, ST · (555) 555-5555 · name@example.com · linkedin.com/in/name
 
SUMMARY
Senior Financial Analyst with 6 years in FP&A. Skilled in Excel,
NetSuite, SQL and 3-statement modelling.
 
EXPERIENCE
Senior Financial Analyst — Company Name, Inc., City, ST | Mar 2022 – Present
- Built the 3-statement model used in the 2024 fundraise.
- Automated reconciliations in NetSuite, cutting close from 9 to 5 days.
 
Financial Analyst — Company Name LLC, City, ST | Jun 2019 – Feb 2022
- Owned monthly variance reporting for a 4-region sales org.
 
EDUCATION
B.S. Finance — University Name, City, ST — 2019
 
SKILLS
Excel · NetSuite · SQL · Power BI · Forecasting · Variance analysis

A few details in it come straight from the vendor guidance:

• Name and contact on the first lines of the body, not in Word's header, because Greenhouse names header and footer contact details as a parse failure.
• Full job titles, because Greenhouse names abbreviated titles. "Senior Financial Analyst," not "Sr. Fin. Analyst."
• Company names with their formal identifier ("Inc.", "LLC") where you know it. Greenhouse's article lists missing formal identifiers among the causes of incomplete parses.
• One line per role carrying title, employer, location and dates, so nothing has to be reassembled from separate columns.
• A plain Skills line, which matters doubly on Workday, where you may also have to type skills into a form field.

The figures in the example are placeholders. Swap in your own, or leave a metric out rather than invent one.

#.docx or PDF?

Both are accepted. Greenhouse's supported formats are ".doc, .docx, .pdf, .rtf, .txt." What matters is that a PDF is text-based: you can select and copy the words. A PDF exported from a design tool as an image, or a scanned page, fails the image test above. The full trade-off is in PDF vs DOCX for ATS.

#Test your format in five minutes

1. Copy-paste test. Open your PDF, select all, and paste into a plain-text editor. If the text comes out in the wrong order, merged across columns, or missing, a parser will see the same.
2. Plain-text save. In Word, use File → Save As → Plain Text and reopen the file. Your name and contact details should be at the top, and each job title should sit next to its own dates.
3. Check the file size. Keep it well under Greenhouse's 2.5MB limit. Text-only resumes are usually tiny.
4. Upload to a real form. Many employer application forms auto-fill from your resume. Blank or wrong fields show you exactly what the parser missed.
5. Run a parse check. Our free resume checker shows the extracted text and flags layout problems, with no account needed.

#What format cannot do

A clean format gets your resume read correctly. It does not make it match. Once the parse is right, the AI compares your qualifications with the job's. HiredScore's release note says basic qualifications decide whether a candidate is graded A/B or C/D. That part is about content: the posting's required skills, titles and years, stated plainly in your experience. For a template that already follows every rule above, see our ATS resume template for Word.

#Sources

• Workday — HiredScore Spotlight Fit and Gap enhancement release note (Aug 2026)
• Workday — Concept: Resume Parsing
• Greenhouse — Unsuccessful resume parse
• Greenhouse — Supported formats for resumes, cover letters and other candidate uploads
• Phenom — Reasons to use AI resume parsing
• Phenom — How to get the most from AI fit scores
• Eightfold — OFCCP AI vendor disclosure (PDF)
• Eightfold Engineering Blog — AI-powered talent matching