ai-job-search: an AI-powered job application framework that automates scraping, tailoring, and ATS-verifying professional LaTeX documents
ai-job-search: an AI-powered job application framework that automates scraping, tailoring, and ATS-verifying professional LaTeX documents
What it solves
Searching for and applying to jobs is often a tedious, manual process involving repetitive cover letter writing and CV tailoring. This project provides an AI-powered framework to automate the discovery, evaluation, and application process while ensuring high-quality, professional output that is actually parseable by Applicant Tracking Systems (ATS).
How it works
Built on Claude Code, the framework uses a structured workflow of specialized commands to manage the job search:
- Profiling:
/setupcreates a candidate profile from existing documents or interviews. - Discovery:
/scrapesearches job portals (with built-in tools for the Danish market and LinkedIn) and/rankscores matches based on fit. - Application:
/applyuses a drafter-reviewer agent pipeline to create tailored LaTeX CVs and cover letters. It includes a unique PDF verification loop that compiles the documents and visually inspects them to fix layout issues (like orphaned titles) and usespdftotextto ensure the PDF text layer is ATS-friendly. - Preparation:
/interviewgenerates stage-specific prep packs and conducts mock interviews based on the candidate's actual experience. - Management:
/outcometracks results, while/notion-syncand/html-reportprovide dashboards for tracking progress.
Who it’s for
Job seekers who want to automate their application pipeline without sacrificing quality or honesty, particularly those comfortable using CLI tools and LaTeX for professional document generation.
Highlights
- Drafter-Reviewer Pipeline: Uses two separate agents to draft and critique applications to avoid generic language.
- PDF Layout Verification: Automatically iterates on LaTeX compilation until the layout is is clean and fits page limits.
- ATS Text-Layer Check: Verifies that the compiled PDF is actually readable by parsers, preventing common LaTeX glyph errors.
- Relevance-Weighted Cutting: Intelligently removes the least relevant experience when a CV exceeds page limits rather than just cutting the oldest entries.
- Comprehensive Career Toolset: Includes skill-gap analysis (
/upskill), company research, and interview preparation.
Sources
- undefinedMadsLorentzen/ai-job-search