AI CV Engine

A JD-aware CV and cover letter generator: one YAML source of truth, NLP keyword scoring against the job description, 12 role-profile classifiers, and a python-docx pipeline that emits polished DOCX/PDF with zero manual editing.

01Why it exists

Tailoring a CV per application is high-value but brutally repetitive: reorder projects, swap emphasis, rewrite the opener, rebuild the PDF. Do it manually and you either burn hours or send generic documents. The engine makes tailoring a build step.

02How it works

1. YAML source of truthThe entire career history lives in structured YAML — experience, projects, skills, metrics — versioned like code.
2. Score against the JDNLP keyword scoring ranks projects and skills by relevance to the specific job description.
3. Classify the role12 role-profile classifiers (LLM / GenAI / Agent / FDE / DS / MLOps / …) each carry a distinct title, summary opener, and cover letter voice.
4. Build the documentspython-docx assembles the tailored CV; LibreOffice converts to PDF. Batch mode processes an entire Excel job tracker in one run.

03Results

• Zero manual editing from YAML to submission-ready PDF
• 12 role profiles mean the same history reads correctly for very different audiences
• Batch mode: tailor for every row of a job tracker in one command
• 30+ pytest assertions keep generated documents structurally correct

04Stack