How ATS Works: Inside Resume Parsing
Understanding how ATS technology works gives you a massive advantage. Here's the technical breakdown of exactly what happens to your resume from upload to recruiter review.
The 5 Stages of ATS Processing
Stage 1: Document Ingestion
When you submit a resume, the ATS first identifies the file format (PDF, DOCX, TXT, RTF). It then extracts raw text using format-specific parsers. PDF parsing uses libraries like Apache PDFBox or Tika. DOCX parsing extracts from the XML structure directly.
What can go wrong: Scanned PDFs (images), encrypted files, corrupted documents, or unusual encodings fail at this stage.
Stage 2: Section Detection
The parser identifies document sections by looking for common headings: "Experience," "Education," "Skills," "Summary." It uses natural language processing and pattern matching to categorize content blocks.
What can go wrong: Non-standard headings ("My Journey," "What I Bring"), multi-column layouts where sections intermix, and tables that break reading order.
Stage 3: Entity Extraction
Named Entity Recognition (NER) models extract structured data:
- Contact info (name, email, phone, LinkedIn)
- Job titles and companies
- Date ranges (employment duration)
- Educational institutions and degrees
- Skills and certifications
Stage 4: Keyword Scoring
The system compares extracted content against the job requirements. This is typically a combination of:
- Exact match: Does the resume contain "Kubernetes"?
- Weighted scoring: "Required" skills worth more than "nice to have"
- Frequency: Keywords appearing multiple times (in context) score slightly higher
- Recency: Keywords in recent roles weighted more than older positions
Stage 5: Ranking & Filtering
Candidates are ranked by composite score. Recruiters typically see only the top 10-25% in their initial view. Some ATS auto-reject candidates below a threshold (usually 60% match).
Common ATS Platforms & How They Differ
| Platform | Market Share | Parsing Quality | Key Note |
|---|---|---|---|
| Workday | ~30% | Good | Requires account creation, strict formatting |
| Greenhouse | ~15% | Excellent | Best PDF parsing, startup-friendly |
| Lever | ~12% | Very Good | Strong NLP, handles creative formats better |
| iCIMS | ~10% | Good | Common in enterprise, conservative parsing |
| Taleo (Oracle) | ~8% | Fair | Legacy system, struggles with modern formats |
See how ATS reads your resume
Upload your resume and see the exact parsing result — what it reads, what it misses, and your match score.
Check my resume freeFrequently Asked Questions
What is an ATS?
An Applicant Tracking System (ATS) is software that companies use to manage job applications. It receives resumes, parses their content into structured data, scores candidates against job requirements, and helps recruiters filter and track applicants through the hiring pipeline.
Which companies use ATS?
99% of Fortune 500 companies and 75%+ of mid-size companies (100-500 employees) use ATS. Popular systems include Workday, Greenhouse, Lever, iCIMS, Taleo, BambooHR, and SmartRecruiters.
Can ATS read images and graphics?
No. ATS can only extract plain text from documents. Images, icons, skill bars, charts, logos, and text embedded in graphics are completely invisible to ATS parsers. Use plain text only.
Does ATS care about resume length?
ATS parses the full document regardless of length. However, keyword density matters — a concise resume with relevant keywords scores better than a long resume padded with irrelevant content. Stick to 1-2 pages.