ATS keyword optimization: populate technical_skills for resume parsing

[adrianwedd]

Context

AI Jobs Australia Resume Keyword Analyser scored the CV at 21% ATS compatibility (17/86 keywords found).

Root Cause

The skills array in base-cv.json had 18 entries but was missing critical ATS keywords: cloud platforms (AWS/Azure/GCP), deep learning, generative AI, MLOps, Kubernetes, and others. The skill naming also used compound names like "LLM Integration" that ATS parsers don't match.

Actions Taken (2026-03-23)

• Added 9 new skills: Cloud Platforms, Kubernetes, Deep Learning, Generative AI, Computer Vision, Data Engineering, MLOps, FastAPI, Bash
• Renamed "LLM Integration" → "Large Language Models" for ATS parsing
• Enriched ML Libraries description with explicit keyword terms (scikit-learn, NumPy, Pandas, NLTK)
• Total skills: 18 → 27

Remaining Work

• Re-run the Resume Keyword Analyser after next pipeline run to measure ATS score improvement
• Consider adding a flat technical_skills array with category groupings for cv-generator.js to render as a dedicated section
• Verify the generated index.html renders all 27 skills prominently

References

• Tool: https://www.aijobsaustralia.com.au/tools/ai-jobs-resume-keyword-analyser
• Skills Gap Analyser scored 79% match against AI Safety Researcher role (3 gaps)