Automated Resume Builder

Manage your professional profile via JSON and get a high-quality LaTeX resume PDF automatically.

Two workflows available:

• CI/CD (Zero Setup): Push to GitHub → PDF auto-generated. No local installation required.
• Local Web UI (Power User): Edit JSON visually, preview PDF in real-time, generate on-demand with full local control.

📄 View Sample Resume

Quick Start

Option A: CI/CD Only (No Local Setup)

1. Use as Template / Fork: Click the "Use this template" button.
2. Edit JSON files in data/.
3. Push to main → GitHub Actions auto-generates PDF.
4. Download from Latest Release.

Option B: Local Web UI (Recommended for Power Users)

1. Install Dependencies:

   uv pip install -r requirements.txt

2. Install TeX Live (for PDF compilation):

   sudo apt install texlive-latex-base texlive-fonts-extra texlive-latex-extra

3. Run the customizer:

   uv run python customizer/server.py

4. Open http://localhost:8000.

Features

• Local Web UI: A minimalist, built-in Customizer UI lets you edit JSON data effortlessly and preview the generated PDF in real-time.
• AI-Tailored Resumes (Multi-Stage Pipeline): Tailor your resume to any job description using a 4-stage pipeline — JD Analysis, Match & Score, Section Tailoring, and Validation. Supports OpenAI, OpenRouter, Cerebras, and Gemini with BYOK. Streams real-time progress via SSE, shows visual diffs, relevance scoring, and evaluation metrics (alignment, content preservation, hallucination detection).
• JSON-based Source of Truth: Manage all your data (profile, experience, education, skills, projects) in structured JSON files.
• LaTeX Professionalism: Utilizes a professional LaTeX template with Jinja2 rendering for a premium look.
• Automated CI/CD: GitHub Actions automatically compiles your LaTeX source into a PDF on every push to main.
• Evaluation Framework: Built-in eval-module with Pydantic schemas, quality metrics (alignment, preservation, hallucination detection), and golden test cases for regression testing.

JSON Data Structure

The data files in data/ directory:

• profile.json: Name, title, bio, and social links.
• experience.json: Professional work history.
• education.json: Academic background.
• skills.json: Categorized technical skills.
• projects.json: Highlighted projects.
• contact.json: Contact information and location.

Local Command Line (Optional)

Generate LaTeX manually without the UI:

python scripts/render_resume.py
pdflatex resume.tex

Accessing Your PDF (CI/CD)

Once you push to GitHub, the CI/CD pipeline auto-generates the PDF:

1. Checking the Latest Release directly.
2. Navigating to the "Releases" section on the right side of your GitHub repository.
3. Downloading the resume.pdf asset from the "Latest" tag.
4. You can check the "Actions" tab to see the build progress and logs.