How ATS Systems Actually Screen Your Resume

By MyCareerAtlas · 8 min · August 26, 2026

The myth: a robot reads your resume and rejects it in milliseconds

The most common piece of job search folklore is some version of "the ATS auto-rejects resumes that don't have the right keywords, before any human ever sees it." It's repeated often enough that it shapes how people write resumes — cramming in keyword variants, avoiding certain formatting, second-guessing every word choice against an imagined algorithm.

The reality is more mundane, and more useful to actually understand.

What an ATS actually is

"ATS" (Applicant Tracking System) refers to the software companies use to manage hiring — posting jobs, collecting applications, and giving recruiters a shared view of every candidate in a pipeline. Greenhouse, Lever, Ashby, and Workday (the platforms MyCareerAtlas tracks) are all, fundamentally, organizational tools for human recruiters, not autonomous rejection engines.

When you apply through one of these, here's what typically happens:

1. Your application (resume, form answers, sometimes a cover letter) is submitted and stored.
2. The system tries to parse your resume — extract your name, contact info, work history, and skills into structured fields, so recruiters can filter and search across candidates.
3. A human recruiter reviews applications, usually sorted or filterable by the parsed fields, sometimes alongside additional criteria they've set up (years of experience, specific required skills).
4. Some companies layer on additional automated screening — a knockout question ("do you have authorization to work in X"), a minimum-requirement filter — but this is a deliberate configuration by that specific employer, not a universal ATS behavior.

The "resume gets auto-rejected by an algorithm before a human sees it" version is real at some large companies with aggressive automated screening configured, but it's far from universal, and it's not what most of these platforms do by default.

The real risk: parsing failure, not keyword rejection

Here's what actually costs candidates interviews, consistently: the ATS's parser misreads your resume, and the recruiter sees an incomplete or garbled version of your information — not because you're unqualified, but because your formatting confused the software extracting it.

Common causes:

• Tables and columns. A two-column resume layout often gets read left-to-right, line-by-line, scrambling the order of your information. What looks clean visually can parse as nonsense.
• Text boxes and headers/footers. Content placed in these often gets skipped entirely by parsers that only read the main document body.
• Unusual fonts or heavy graphic design. Icons, non-standard fonts, and heavy visual styling can interfere with text extraction, even when the underlying document looks great to a human.
• Images of text. If any part of your resume is an image (a designed header, a logo with your name in it) rather than actual text, that content is often invisible to the parser entirely.

None of this is about "beating an algorithm" with the right keywords. It's about making sure the software correctly extracts what's actually true about your background, so the human reviewing it sees accurate information instead of a mangled version of it.

Practical formatting advice that actually holds up

• Standard, single-column layout. Reverse-chronological, one column, clear section headers (Experience, Education, Skills). Not exciting, but reliably parseable across every major ATS.
• Real text, not images. Every word should be selectable text, including your name and contact info at the top.
• Standard section headers. "Work Experience" or "Professional Experience" parses more reliably than creative alternatives like "My Journey."
• Save as a standard file type. PDF (text-based, not a scanned image) or .docx both parse reliably on modern platforms. Avoid unusual formats.
• Include keywords naturally, not as a separate block. If a posting emphasizes a specific skill you genuinely have, use the same terminology the posting uses somewhere in your actual experience descriptions — not because an algorithm is scanning for it, but because it makes it easy for a human skimming quickly to confirm the match.

Where the platforms genuinely differ

The core advice above holds across Greenhouse, Lever, Ashby, and Workday, but there are practical differences worth knowing:

• Greenhouse and Lever are widely used by startups and mid-size tech companies, tend to have relatively standard, clean application forms, and are generally reliable at parsing well-formatted resumes.
• Ashby is newer and used heavily by fast-growing tech companies; its application flows are often similarly clean, sometimes with more custom screening questions per role.
• Workday is common at large enterprises, has a reputation (deserved, in practice) for longer, more form-heavy application flows with more required fields — budget more time per application, and expect to manually re-enter information a smaller-company ATS might have parsed automatically.

The actual takeaway

Applicant tracking systems are workflow tools for recruiters, not the adversarial gatekeeper internet folklore describes. The real risk isn't a keyword algorithm rejecting you — it's formatting choices that cause honest parsing errors, which is a fixable, mechanical problem, not a "figure out what the robot wants" guessing game.

MyCareerAtlas discovers roles directly from Greenhouse, Lever, Ashby, and Workday boards and shows you a match score with the actual reasoning behind it — not a black-box percentage.