ATS Resume Score Example: Reproduce 76 to 88 With Text and PDF
The fixed JD contains five equally weighted skill terms. The fictional resume initially matches three: 60% keyword coverage and 100% text checks produce 76. Naming Kubernetes, a skill assumed to be real in this fictional scenario, matches four terms: 80% coverage produces 88.
What does this example demonstrate?
This is a deterministic keyword and text-extraction demonstration, not a real applicant, customer case, hiring experiment or interview-improvement claim. The JD is a controlled five-term input, not a full job posting. The resume is synthetic; do not copy its added skill into your own experience.
Download the inputs and outputs
- Fixed JD
- Before resume text
- After resume text
- Before sample PDF
- After sample PDF
- Before PDF extracted text
- After PDF extracted text
- Expected results JSON
These PDFs are static demonstration attachments, not customer documents exported through the paid studio. Text was extracted with PyMuPDF and checked with the same scoring functions used by the website. Other PDF parsers may produce different whitespace or reading order.
Verified calculation
| Input | Matched terms | Coverage | Text checks | Simulated score |
|---|---|---|---|---|
| Before text / PDF | Python, SQL, Docker | 60% | 100% | 76 |
| After text / PDF | Python, SQL, Docker, Kubernetes | 80% | 100% | 88 |
round(0.60 × 60 + 0.40 × 100) = 76round(0.60 × 80 + 0.40 × 100) = 88
How can you reproduce it?
- Paste the fixed JD and before text into the matching homepage fields. Run only the local check; no AI call is needed.
- Check that the selected terms are Python, SQL, Docker, Kubernetes and AWS, each appearing once in the JD.
- Replace the resume with the after text and check again. Only Kubernetes is added; AWS remains absent.
- Upload each sample PDF, or paste its provided extracted text, and compare keyword coverage and text checks.
- If your result differs, inspect the selected terms and dictionary loading, then dates, length and copied formatting. This example must be rechecked when scoring rules change.
Why is this not a keyword-stuffing recipe?
The example isolates how literal matching changes when an existing skill is named explicitly. It does not demonstrate better experience quality. Repeating a matched term does not improve coverage; AWS remains missing because the fictional evidence does not support it. AI suggestions and high scores cannot replace real experience.
Method and version
Read the scoring formula, text-check rules and data handling · Source repository
Verified November 7, 2026 using the current production scoring functions and skills dictionary; not an employer ATS benchmark.