By Peter Miller · Published August 10, 2026 · Last updated: September 15, 2026
11 min read · You don't need a template swap tonight
Your PDF shows Kubernetes migrations and Python services. Greenhouse shows a jumbled paragraph and the wrong employer on bullet two. That's not a hidden score penalty. It's export layout. Fintech, SaaS, and enterprise IT teams in London, Manchester, and Edinburgh run Greenhouse on backend, platform, and DevOps reqs. When you park Month Year dates in a table cell or list cloud tools only under Skills, the parser flattens blocks before a human opens the file.
Before you rewrite bullet verbs, check your resume for free with the posting pasted in. Paste your export into Notepad first. If employer names and dates merged into one string, structure broke upstream. You won't rescue an IT upload by stuffing AWS terms into a Skills footer.
You've probably been told to quantify uptime and release cadence. Good advice, useless when bullet three never imported as its own line. Job searching through corporate portals is already slow, especially when you're mid-level and competing on stack depth, not titles alone. This diagnostic walks symptom to cause to fix. You don't need a template swap tonight.
And it is not only junior applicants. A mid-level engineer with six years of platform work can miss the screen because bullet one still says supported deployments while the posting's lead requirement was Kubernetes on AWS. The years are real. The parsed view just never attached them to the right employer block.
If the portal asks for a cover letter, don't paste your Skills block into paragraph one. Generate a cover letter that names the same stack you put on bullet one, then stop rewriting.
Do these three things first
- Do not park Month Year dates in table cells. Put them on the same line as job title and company.
- Use one date pattern across the whole file: Jan 2022 to Present, not mixed with 01/2022.
- Name Python, AWS, or Kubernetes in bullet one under the employer where you used them, not only in Skills.
The silence when your UK IT CV parses wrong in Greenhouse
The symptom is quiet. Portal status stays submitted. No rejection email. You open the file and the Skills section reads like a copy of the job description. So you assume the problem is volume or bad luck.
Parsing failures often show up before a human ever reads bullet one.
Greenhouse attaches your upload to a candidate profile tech recruiters skim inside the platform. They see parsed text alongside or instead of your PDF. For software engineer, platform, and DevOps roles at mid level, the first ctrl-f is often the cloud stack, then language, then release scope. When import flattens structure, those fields disappear or attach to the wrong employer line.
A composite backend engineer whose PDF shows dates in a right-column cell but Greenhouse imports one jumbled block loses to the same facts on plain single-column lines. Same qualifications. Different skim surface for the coordinator forwarding to an engineering manager.
For how parsers treat Skills differently from employer bullets, see how ATS interprets skills vs tools vs technologies . The short version: tags are not proof.
Greenhouse ATS resume optimization UK IT CVs: three parsing causes
Diagnostic posts fail when they list twelve fixes. You need one decision: which layer broke your file. Work through the three causes below in order. Stop when you find a clear miss.
Cause 1: Two-column layouts and date tables scramble employer blocks
This is the most common break on UK IT CVs built from Canva or Word templates with a sidebar. Job title and company sit in column one. Month Year dates sit in a right-column table cell. Greenhouse reads top to bottom inside the file, not left to right across your design.
Before: Two-column template. Left column: Software Engineer, Northwind Digital Ltd, London. Right column table cell: Jan 2022 to Present. Bullets below in a shaded text box.
After: Single-column plain block:
Software Engineer | Northwind Digital Ltd | London | Jan 2022 to Present
Migrated 14 microservices to Kubernetes on AWS EKS; cut deploy failures 22% in H1 2025.
I have screened Greenhouse stacks where a platform engineer's Kubernetes bullets buried inside a skills table imported after job two, so the parser attached EKS work to a six-month contract from 2021. The recruiter never reached the uptime scope because the parsed view looked like job hopping without stack proof up top.
Cause 2: Mixed UK date formats break tenure anchors
Cause two looks like cause one but the layout is already single-column. The dates themselves fight each other. One role uses Jan 2022 to Present. The next uses 01/2022 to 12/2023. A contract row says Q3 2021. Greenhouse cannot anchor tenure when the pattern shifts mid-file.
Before: DevOps Engineer, Apex Retail Group, 03/2019 to 08/2022. Next role: Backend Engineer, Jan 2022 to now. Third role: Contract SRE, Q1 2021.
After: DevOps Engineer, Apex Retail Group, Mar 2019 to Aug 2022. Backend Engineer, Horizon Media Ltd, Jan 2022 to Present. Contract SRE, Sterling Logistics, Jan 2021 to Mar 2021.
Edge case: contract and permanent mix at mid level. List each engagement as its own employer block with Month Year start and end. Do not fold three contracts into one line with commas.
Cause 3: Skills holds every keyword but experience bullets stay generic
Cause three looks like a wording problem but the fix is placement. You named the right tools in Skills and repeated them in a summary paragraph. None of them sit on a dated employer line with a metric.
Before: Skills: Python, AWS, Kubernetes, Docker, Terraform, CI/CD, REST APIs. Experience bullets: Worked on backend features. Supported deployments. Collaborated with cross-functional teams.
After: Software Engineer, Cedar Labs, Jun 2021 to Present. Built Python REST services on AWS ECS; cut p95 latency 18% in Q3 2025. Owned Terraform modules for 3 production environments.
Edge case: you used the tool five years ago, not in your current role. Put it on the older job with Month Year dates. Do not float it only in Skills for a posting that wants recent use.
How to tell which cause is yours
- Layout check: Paste your export into Notepad. If employer names and Month Year digits merged into one string, you are in Cause 1.
- Date check: Scan every employer line. If you see 01/2022 beside Jan 2022 beside Q1 2021, you are in Cause 2.
- Skills check: ctrl-f Python, AWS, or the posting's top stack in the Notepad paste. If hits appear only in a Skills block, you are in Cause 3.
Read how to tell if your resume is ATS friendly in 60 seconds when you want a faster layout pass before you open Word again.
Copy-paste employer block (edit every bracket)
[Job Title] | [Company Ltd] | [City] | [Mon Year] to Present
• [Verb] [Python/AWS/K8s scope] for [object]; [metric] in [timeframe]
• [Second bullet only if it supports the same stack]
• [Third bullet optional; do not bury the must-have here]
Skills (short; each tool above must appear on a bullet):
[capability] · [capability] · [cloud stack already on bullet one]
Copy-paste this block, then replace every bracket with your real stack and dates.
Where UK IT files still break on Greenhouse import
Even after you fix the cause that blocked you, these habits put the next apply back in the maybe pile.
Headers and footers with contact details or GitHub links. Greenhouse often drops header and footer text on import. If your phone number or portfolio URL lived only there, the parsed profile looks incomplete.
Icons and emoji for email or phone. Parsers read them as blank characters. Write Phone: and Email: as plain labels on line one of the file.
Creative job titles like Code Ninja instead of Software Engineer. Recruiters ctrl-f for standard titles. Parsers keyword-match the same strings.
Chasing keyword density in Skills while bullet one still says worked on features. Shipped, built, cut, or migrated with a number tells the desk what you did last quarter.
Uploading a designed PDF from a portfolio site when the portal accepts DOCX. Word exports with standard heading styles usually keep employer blocks separated better than Canva PDFs with invisible tables.
Saving the file with a generic name like CV.pdf instead of Firstname_Lastname_SoftwareEngineer.docx. Recruiters search folders by role title. A clear filename helps humans even when parsers ignore it.
Parse check before you reapply
Rebuilt layout still fails if the checker shows your must-have only in Skills. Upload the file you will send and run a free ATS check against the same posting. Confirm Python, AWS, or the posting's top term appears inside extracted experience text, not only in a footer row.
Fix layout and dates first. A longer Skills list cannot invent proof on a dated line.
When you are choosing which reqs deserve a rewrite tonight, score your job match on the posting before you spend an hour on bullets three and four.
Rebuild plain structure tonight
That is the file a desk can forward. Skills still matter as tags. Bullet one under your current employer is where stack proof lives. When those disagree, you stay in the maybe pile even with the right qualifications on paper.
Open the posting. Highlight the first requirement. Rebuild one single-column employer block with Month Year on the title line. Rewrite bullet one so the first eight words answer the posting with a language, cloud service, and metric.
Job searching is already hard. You do not need a longer keyword list. You need a plain export Greenhouse can read. When you are ready to verify the wording, check your resume for free against the same posting before you send it.
Frequently asked questions
Greenhouse reads a text stream, not your table layout. When job titles sit in column one and Month Year dates sit in a right-column cell, the parser often imports dates after the company line. Bullet blocks then slide onto the previous or next employer. Move dates onto the same plain line as title and company, then put cloud stack scope in bullet one under that employer block.
Use Month Year consistently: Jan 2022 to Present or Sep 2019 to Mar 2023. Avoid mixing 01/2022, Q1 2022, and spelled-out January on the same file. Parsers struggle when tenure math cannot anchor to one pattern. Put the range on the employer line, not in a footer, header, or sidebar cell.
Try DOCX first when the portal allows it. Word files with standard heading styles usually keep employer blocks separated better than designed PDFs from template sites. If PDF is required, export from Word after you strip tables and columns. Paste the export into Notepad before you apply so you see what the recruiter view receives.
No. Skills rows store tags. Greenhouse and the recruiter both look for dated proof under employer titles. If Python, AWS, and Kubernetes sit only in Skills while bullets read supported deployments and worked on features, the parsed profile looks thin even when the checker scores well. Move each must-have from Skills into bullet one on the job where you used it.
Yes when the portal lets you replace the file or a recruiter requests an updated CV. Rebuild with plain employer lines, consistent Month Year dates, and stack scope in bullet one. Paste into Notepad to confirm each bullet starts on its own row, then upload once. Note in one sentence that you fixed formatting so scope parses cleanly.