Resume Reworker 📄✨

An AI-powered application to analyze, rewrite, and instantly generate ATS-friendly LaTeX resumes tailored to any job description.

🎯 The Idea

The days of manually tweaking your resume for every job application are over. Resume Reworker is an intelligent agent that takes your base resume and a target job description, and handles the rest:

1. Analyzes your resume against the job description to find missing keywords and weak points.
2. Rewrites your bullet points to naturally incorporate keywords and highlight the most relevant skills—without hallucinating or fabricating experience.
3. Generates a beautifully typeset, ATS-compliant PDF on the fly using a dedicated LaTeX compilation engine.

✨ Features

• Smart ATS Analysis: Get immediate feedback on what keywords you missed and how to improve.
• Agentic Rewriting: Powered by LangGraph, the AI pipeline breaks down rewriting into robust, reliable steps that avoid data corruption (e.g., losing hyperlinks or contact info).
• Multi-Model Support: Plug in your API keys for Google Gemini, Mistral, Groq, or HuggingFace models.
• Live PDF Preview: See your LaTeX resume compile instantly as you tweak the results.
• Built-in Templates: Choose from industry-standard LaTeX templates (like Jake's Resume Template).
• Interactive UI: Built for speed with a modern React + Vite + Tailwind interface.

---

🏗️ Architecture

graph TD
    A[User UI - React/Vite] -->|Upload PDF & JD| B(FastAPI Backend)
    B --> C{LangGraph Workflow}
    
    C -->|Step 1| D[Extract Details & Links]
    C -->|Step 2| E[Analyze against JD]
    C -->|Step 3| F[Rewrite Bullets & Skills]
    
    F --> B
    B --> G[LaTeX Renderer]
    G --> H[(Docker LaTeX Compiler)]
    H -->|Compiled PDF| A

Loading

🛠️ Tech Stack

Frontend

• Framework: React 19 + Vite
• Styling: Tailwind CSS v4 + Framer Motion
• State Management: Zustand + SWR
• Components: Shadcn UI & Radix Primitives

Backend

• API: FastAPI (Python 3.14)
• Database: PostgreSQL (SQLAlchemy + asyncpg)
• Caching & Rate Limiting: Redis
• AI Orchestration: LangGraph + LangChain

Infrastructure

• PDF Generation: Custom isolated Docker container (latex-compiler)
• Containerization: Docker Compose

---

🚀 Setup & Installation

Prerequisites

1. Docker & Docker Compose (required for the database, Redis, and LaTeX compilation engine).
2. Node.js (v20+ recommended) for the frontend.
3. Python 3.14+ for the backend.

1. Backend Setup

cd backend
# Create a virtual environment
python -m venv .venv
source .venv/bin/activate

# Install dependencies
pip install -r requirements.txt

# Start the infrastructure (PostgreSQL, Redis, LaTeX compiler)
docker-compose up -d

# Set up environment variables
cp .env.example .env
# Edit .env with your DB credentials and JWT secrets

# Run the backend
./backend.sh

2. Frontend Setup

cd frontend
# Install dependencies
npm install

# Run the development server
npm run dev
# OR use the provided script
./frontend.sh

3. Usage

• Navigate to http://localhost:5173
• Create an account or sign in with Google OAuth.
• Navigate to the Profile section to input your API keys (e.g., Gemini, Groq).
• Upload your current resume PDF and paste a Job Description to begin the AI workflow!

---

📝 Configuration

• Prompt Engineering: You can adjust the system prompts and strict extraction rules in backend/utils/prompts.py.
• LaTeX Templates: The base LaTeX templates are stored as Jinja2 templates in backend/templates/. You can customize spacing, fonts, or colors here.

🤝 Contributing

Contributions are welcome! Please open an issue or submit a pull request if you want to add new LaTeX templates or support for new LLM providers.