LinkedIn PDF → JSON Resume / Europass XML / LaTeX / SWE PDF

Keep a single source of truth for your resume and export it from a LinkedIn PDF into JSON Resume, Europass XML, and a clean SWE LaTeX/PDF CV. You can customize the output via supplying additional info or editing intermediate outputs. For best experience also export LinkedIn data archive from Settings&Privacy to supply the full info to the output. No LLM required. Built via OpenAI Codex.

What this tool can do

• Parse a LinkedIn profile PDF into JSON Resume (resume.json) using the open schema: https://jsonresume.org
• JSON Resume can be converted into many community themes; explore options at https://jsonresume.org/themes and the CLI at https://github.com/jsonresume/resume-cli.
• Merge extra personal details from personal_info.json (e.g., phone, birth date, nationality, additional_skills) where LinkedIn lacks data.
• Export JSON Resume to Europass Candidate XML (europass.xml) for import into the Europass editor: https://europass.europa.eu/en
• Render JSON Resume into a customized version of Jake's LaTeX resume template (resume.tex) and compile a PDF (resume.pdf), including optional sections like summary, certificates, and languages. A base template is also available as template_basic.tex.

Example input and output

LinkedIn export input (first page) and LaTeX SWE output (first page):

Disclaimer

LinkedIn PDF parsing is best-effort and may miss or misclassify fields. Always review the generated resume.json, europass.xml, and resume.pdf before using them. If you want to tailor the output to your resume style or data, use AI in this repo to adjust the underlying Python code. Use malformed resume.json output and any missing fields to pinpoint what needs fixing.

Background

I keep my LinkedIn profile up to date and wanted a single source of truth that could generate multiple formats: Europass, and a clean PDF using Jake's popular SWE resume template. I then modified Jake's template to include more of LinkedIn profile exports.

Why not a web app?

A hosted web app would need robust PDF parsing and résumé cleanup, which typically means an LLM in the loop. That turns it into a paid service (like sweresume.app 5$ per generation). This repo keeps everything local and scriptable, so if you are an engineer (especially with a Copilot license, personal or work-related), you can clone it and tailor the output quickly. If it helps, all I ask in return is a star on the repo.

Template notes

The LaTeX template in template.tex is a modification of Jake's original resume layout. It adds extra sections (summary, certifications, languages) and a compact contact/header layout to fit typical LinkedIn exports.

Requirements

• Python 3.11+
• LaTeX engine: pdflatex for ASCII-only output, xelatex or lualatex for Unicode output (MiKTeX or TeX Live). MiKTeX download: https://miktex.org/download

Installation

pip install -r requirements.txt

Usage

LinkedIn export: open the profile, click More → Save to PDF.

Quick start (full pipeline): Fill out personal_info.json first (phone, location, nationality, additional_skills, etc.) so missing LinkedIn data is merged into the output. If you have LinkedIn exports for skills, certifications, or projects, drop them as Skills.csv, Certifications.csv, and Projects.csv to merge into the resume data. LinkedIn export path: Settings & Privacy → Data privacy → Get a copy of your data → Request archive (All data).

.\linkedin-to-latex-pdf.ps1 -LinkedInPdf path\to\LinkedIn_Profile.pdf

Outputs are named from the input PDF (e.g., LinkedIn_Profile_resume.json, LinkedIn_Profile_europass.xml, LinkedIn_Profile_SWE.pdf). Add -Basic to keep only the original template sections in the LaTeX output. Use -Latinize to transliterate non-ASCII text and force pdflatex, or -ForcePdfLatex to skip Unicode engine switching. The LaTeX output defaults to a light theme; use -Dark for a dark background.

Optional arguments (defaults)

Only the LinkedIn PDF path is required; everything else has a sensible default.

linkedin-to-latex-pdf.ps1:

Argument	Required	Default
-LinkedInPdf	Yes	(none)
-ResumeJson	No	<pdf>_resume.json
-PersonalInfo	No	personal_info.json
-SkillsCsv	No	Skills.csv
-CertificationsCsv	No	Certifications.csv
-ProjectsCsv	No	Projects.csv
-Template	No	template.tex
-TexOut	No	<pdf>_SWE.tex
-EuropassXml	No	<pdf>_europass.xml
-EuropassConfig	No	europass_config.json
-SkipEuropass	No	false
-Basic	No	false
-Latinize	No	false
-ForcePdfLatex	No	false
-Dark	No	false

json-to-latex-pdf.ps1:

Argument	Required	Default
-Resume	No	resume.json
-Template	No	template.tex
-TexOut	No	resume.tex
-Basic	No	false
-Latinize	No	false
-ForcePdfLatex	No	false
-Dark	No	false

Python modules:

Command	Required	Defaults
python -m linkedin_resume_parser.cli <pdf>	<pdf>	-o resume.json --personal-info personal_info.json --skills-csv Skills.csv --certifications-csv Certifications.csv --projects-csv Projects.csv
python -m linkedin_resume_parser.europass [resume.json]	(none)	resume.json -m personal_info.json -c europass_config.json -o europass.xml
python -m linkedin_resume_parser.latex [resume.json]	(none)	resume.json -t template.tex -o resume.tex

Individual steps: Parse a LinkedIn PDF into resume.json:

python -m linkedin_resume_parser.cli path\to\LinkedIn_Profile.pdf -o resume.json

Export resume.json to Europass Candidate XML (defaults to resume.json if omitted):

python -m linkedin_resume_parser.europass -m personal_info.json -c europass_config.json -o europass.xml

Render resume.json to LaTeX and PDF (defaults to resume.json if omitted; uses xelatex/lualatex when Unicode is detected):

.\json-to-latex-pdf.ps1 -Template template.tex -TexOut resume.tex

You can edit the generated resume.tex to polish content or formatting before compiling the PDF. Basic (original template sections only):

.\json-to-latex-pdf.ps1 -Template template_basic.tex -TexOut resume_basic.tex

The same flag works with the Python renderer; it will auto-select template_basic.tex when you keep the default template.tex:

python -m linkedin_resume_parser.latex -t template.tex -o resume.tex --basic

Add --latinize to transliterate non-ASCII text and keep the pdflatex pipeline. Use --font "Font Name" to override the Unicode font. Add --dark for a dark theme.

Files

• *_resume.json: JSON Resume output from the LinkedIn parser.
• personal_info.json: Personal metadata not exported from LinkedIn (e.g., birth date, nationality, additional_skills).
• Skills.csv: Optional LinkedIn skills export (Name column) that is merged into resume.json. Get it from LinkedIn: Settings & Privacy → Data privacy → Get a copy of your data → Request archive (All data).
• Certifications.csv: Optional LinkedIn certifications export used to enrich resume.json certificates (issuer, dates, URLs).
• Projects.csv: Optional LinkedIn projects export used to enrich resume.json projects (description, dates, URLs).
• europass_config.json: Europass technical configuration (schemes, rendering, section order, ESCO URI).
• *_europass.xml: Europass Candidate XML output.
• *_SWE.tex: LaTeX output from the resume renderer.
• *_SWE.pdf: PDF output from the selected LaTeX engine (pdflatex, xelatex, or lualatex).

TODO

• Explore LinkedIn alternatives (hh.ru?)
• Write a LinkedIn post about the project + add to my CV projects