📄 Resume Optimization

Resume Optimization is an end-to-end, LLM-powered pipeline that transforms a .docx resume into a tailored, ATS-friendly PDF aligned with a specific job description. It leverages OpenAI or Gemini for semantic rewriting and includes a visual HTML editor for layout polishing.

---

🚀 What It Does

• 🧾 Converts .docx resumes into structured Markdown
• 🧠 Uses LLMs (GPT/Gemini) to adapt the content to any job ad
• 🧰 Applies intelligent style mapping and formatting
• 🖋️ Allows visual editing (PyQt5 WYSIWYG)
• 📄 Outputs a polished, print-ready PDF

---

📁 Project Structure

Resume-Optimization/
│
├── assets/fonts/                 # Custom font files (Garamond, Source Sans)
├── cv_template/                  # DOCX styling template (Word)
├── docs/                         # Markdown documentation
├── logs/                         # Runtime logs
├── original_docx/                # Input resumes (.docx)
├── processed_cv/                 # Intermediate files (.md, .html, .txt)
├── pdf_cv/                       # Final exported resumes (.pdf)
│
├── src/                          # Python source code
│   ├── convert_to_md.py          # DOCX → Markdown
│   ├── optimize_resume.py        # Build prompt
│   ├── adapt_resume.py           # Generate adapted Markdown
│   ├── export_resume.py          # HTML generation + visual editor + PDF export
│   └── __init__.py
│
├── main.py                       # 🔁 Orchestrates full ETL pipeline
├── requirements.txt              # Pip dependencies
├── environment.yml               # Conda environment (optional)
├── .env / .env.example           # API keys and configuration
├── README.md
├── CHECKLIST.md                  # Manual QA/testing checklist
└── CHANGELOG.md

---

🧠 How It Works

Just run:

python main.py

This will:

1. ✅ Extract and map styles from original_docx/
2. ✍️ Adapt the content using GPT or Gemini based on job_description.txt
3. 🧱 Rebuild the resume structure semantically
4. 🎨 Open an interactive WYSIWYG HTML editor for final tweaks
5. 📄 Export a clean, custom-styled PDF in pdf_cv/

No need to run individual scripts manually — the full process is handled by main.py.

---

🗺️ Pipeline Architecture

---

🛠 Requirements

• Python 3.10+
• wkhtmltopdf installed and added to PATH
• Fonts:
  • Garamond Premier Pro
  • Source Sans 3 (Regular)

Install dependencies:

pip install -r requirements.txt

Or use Conda:

conda env create -f environment.yml
conda activate resume_optimization

---

🔐 Setup

1. Copy .env.example → .env
2. Add your keys:

OPENAI_API_KEY=your_openai_key
GOOGLE_API_KEY=your_google_key

---

📌 Features

• Full resume pipeline: .docx → Markdown → Adaptation → HTML → PDF
• Job-specific rewriting using GPT or Gemini
• Visual HTML editor (PyQt5)
• Custom typographic styling (Garamond, Source Sans)
• Controlled spacing (line-height, bullets, paragraphs)
• PDF preview auto-open
• Structured logging and modular code

---

👨‍💻 Author

Created by Manuel Cruz Rodríguez
Humanities graduate pivoting into Data Science & NLP.

🔗 LinkedIn
📬 Feedback, issues and suggestions welcome.

---

📘 License

MIT License.
Feel free to fork, adapt and improve — just don't ship Comic Sans. 😉