Full-Stack AI-Powered Resume Optimization Tool
Built with FastAPI, LLM (OpenAI/Gemini), Vector Search, and Modern Web Stack
This is a production-ready, recruiter-impressive ATS (Applicant Tracking System) resume generator that:
- Extracts job-specific keywords from job descriptions using LLMs (OpenAI/Gemini)
- Generates ATS-optimized resumes in Word (.docx) format with intelligent keyword injection
- Provides a modern web UI for recruiters and job seekers
- Uses vector embeddings to match candidate experience with job requirements (RAG)
- Maintains 1-page C3 resume format for maximum ATS compatibility
Perfect for showcasing: Full-Stack Development, ML/LLM Engineering, and Production-Ready Software Architecture.
┌─────────────────────────────────────────────────────────────────┐
│ Client Layer │
│ ┌──────────────┐ ┌──────────────┐ ┌─────────────────────┐ │
│ │ Web Browser │ │ Mobile UI │ │ API Client │ │
│ │ (HTML/JS) │ │ (Future) │ │ (cURL/Postman) │ │
│ └──────┬───────┘ └──────┬───────┘ └──────────┬──────────┘ │
└─────────┼──────────────────┼──────────────────────┼────────────┘
│ │ │
│ HTTP/REST │ │
└──────────────────┼──────────────────────┘
│
┌────────────────────────────┼─────────────────────────────────────┐
│ FastAPI Application Layer │
│ ┌──────────────────────────────────────────────────────────┐ │
│ │ FastAPI Server (Uvicorn) │ │
│ │ ┌─────────────┐ ┌──────────────┐ ┌──────────────┐ │ │
│ │ │ Routes │ │ Middleware │ │ Auth/JWT │ │ │
│ │ │ /generate │ │ (CORS, │ │ (Future) │ │ │
│ │ │ /download │ │ Logging) │ │ │ │ │
│ │ │ /health │ │ │ │ │ │ │
│ │ └──────┬──────┘ └──────┬───────┘ └──────┬───────┘ │ │
│ └─────────┼─────────────────┼──────────────────┼──────────┘ │
└────────────┼─────────────────┼──────────────────┼─────────────┘
│ │ │
┌────────────┼─────────────────┼──────────────────┼─────────────┐
│ ┌───────▼───────┐ ┌──────▼───────┐ ┌──────▼──────┐ │
│ │ Business │ │ LLM Client │ │ Vector │ │
│ │ Logic │ │ (OpenAI/ │ │ Search │ │
│ │ Layer │ │ Gemini) │ │ (RAG) │ │
│ │ │ │ │ │ │ │
│ │ • Resume │ │ • Keyword │ │ • Embed │ │
│ │ Generator │ │ Extract │ │ JD │ │
│ │ • Template │ │ • Bullet │ │ • Match │ │
│ │ Engine │ │ Rewrite │ │ Exp. │ │
│ │ • Formatting │ │ • STAR │ │ • Retrieve │ │
│ └───────┬───────┘ └──────┬───────┘ └──────┬──────┘ │
│ │ │ │ │
┌────────────┼──────────────────┼─────────────────┼─────────────┐
│ ┌───────▼────────┐ ┌──────▼──────┐ ┌──────▼──────┐ │
│ │ Data Layer │ │ Storage │ │ External │ │
│ │ │ │ │ │ Services │ │
│ │ • PostgreSQL │ │ • Vector │ │ │ │
│ │ (Future) │ │ DB │ │ • OpenAI │ │
│ │ • Pydantic │ │ (Chroma/ │ │ API │ │
│ │ Models │ │ Qdrant) │ │ • Gemini │ │
│ │ • File System │ │ │ │ API │ │
│ └────────────────┘ └─────────────┘ └─────────────┘ │
└──────────────────────────────────────────────────────────────┘
┌──────────────────────────────────────────────────────────────┐
│ USER INPUT │
│ 1. Paste Job Description (JD) │
│ 2. Optionally: Upload existing resume / experience data │
└──────────────────────┬───────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────────────┐
│ STEP 1: JOB DESCRIPTION PROCESSING │
│ ┌──────────────────────────────────────────────────────┐ │
│ │ • Validate JD format (Pydantic) │ │
│ │ • Preprocess text (clean, normalize) │ │
│ │ • Extract metadata (title, company, location) │ │
│ └────────────────────┬──────────────────────────────────┘ │
└───────────────────────┼──────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────────────┐
│ STEP 2: LLM KEYWORD EXTRACTION │
│ ┌──────────────────────────────────────────────────────┐ │
│ │ OpenAI/Gemini API Call: │ │
│ │ • Prompt: "Extract 20-40 skills/keywords from JD" │ │
│ │ • Model: gpt-4o-mini / gemini-pro │ │
│ │ • Output: JSON array of keywords │ │
│ └────────────────────┬──────────────────────────────────┘ │
└───────────────────────┼──────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────────────┐
│ STEP 3: VECTOR SEARCH (RAG) - OPTIONAL │
│ ┌──────────────────────────────────────────────────────┐ │
│ │ IF user has uploaded experience data: │ │
│ │ • Embed JD keywords using sentence-transformers │ │
│ │ • Query vector DB for matching experience bullets │ │
│ │ • Retrieve top 5-10 most relevant experience items │ │
│ └────────────────────┬──────────────────────────────────┘ │
└───────────────────────┼──────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────────────┐
│ STEP 4: RESUME GENERATION │
│ ┌──────────────────────────────────────────────────────┐ │
│ │ • Load template (c3_template.docx) or create new │ │
│ │ • Inject keywords into "Key Skills" section │ │
│ │ • Format with python-docx (1-page limit) │ │
│ │ • Apply ATS-friendly formatting │ │
│ └────────────────────┬──────────────────────────────────┘ │
└───────────────────────┼──────────────────────────────────────┘
│
▼
┌──────────────────────────────────────────────────────────────┐
│ STEP 5: OUTPUT │
│ • Save DOCX to outputs/ directory │
│ • Return file download link or direct file response │
│ • Log generation metadata (for analytics) │
└──────────────────────────────────────────────────────────────┘
| Layer | Technology | Purpose |
|---|---|---|
| Web Framework | FastAPI | High-performance async API + auto-generated docs |
| LLM Provider | OpenAI / Google Gemini | Keyword extraction, bullet rewriting, STAR formatting |
| Document Generation | python-docx | Word (.docx) resume creation with precise formatting |
| Vector Search | sentence-transformers + Chroma/Qdrant | RAG for matching experience with job requirements |
| Validation | Pydantic v2 | Type-safe API request/response validation |
| Templating | Jinja2 | Dynamic HTML generation for web UI |
| Technology | Purpose |
|---|---|
| HTML5 + Tailwind CSS | Modern, responsive web UI |
| Vanilla JS / Alpine.js | Lightweight interactivity (no heavy frameworks) |
| Fetch API | Async API calls to FastAPI backend |
| Technology | Purpose |
|---|---|
| Docker | Containerization for portable deployment |
| Docker Compose | Multi-service orchestration (app + DB + vector DB) |
| GitHub Actions | CI/CD pipeline (tests → build → deploy) |
| PostgreSQL (Future) | Persistent storage for user accounts, job history |
| Prometheus / Grafana (Future) | Metrics and observability |
| Technology | Purpose |
|---|---|
| OpenAI API | GPT-4o-mini for cost-effective keyword extraction |
| sentence-transformers | Generate embeddings for JD and experience bullets |
| Chroma / Qdrant | Vector database for semantic search |
ats_resume_app/
│
├── README.md # This file
├── requirements.txt # Python dependencies
├── .gitignore # Git ignore rules
├── Dockerfile # Docker container config
├── docker-compose.yml # Multi-service orchestration (Future)
│
├── resume_templates/ # Word template files
│ └── c3_template.docx # C3 format resume template
│
├── outputs/ # Generated resumes (gitignored)
│ └── .gitkeep
│
├── src/ # Application source code
│ ├── __init__.py
│ ├── main.py # FastAPI app entry point
│ ├── models.py # Pydantic models for validation
│ ├── resume_generator.py # Core Word generation logic
│ ├── llm_client.py # OpenAI/Gemini API integration
│ ├── vector_search.py # RAG/vector search (Future)
│ └── utils.py # Helper functions
│
├── static/ # Static assets
│ └── styles.css # CSS styles
│
├── templates/ # HTML templates
│ └── index.html # Main web UI
│
├── tests/ # Test suite (Future)
│ ├── test_resume_generator.py
│ ├── test_llm_client.py
│ └── conftest.py
│
└── .github/ # CI/CD workflows
└── workflows/
└── ci.yml # GitHub Actions pipeline
- Python 3.11+
- pip
- (Optional) Docker + Docker Compose
- Clone and navigate to the project:
cd ats_resume_app- Create a virtual environment:
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate- Install dependencies:
pip install -r requirements.txt- Set environment variables:
# On Windows PowerShell:
$env:OPENAI_API_KEY = "sk-your-key-here"
# On Linux/Mac:
export OPENAI_API_KEY="sk-your-key-here"- Run the application:
uvicorn src.main:app --reload- Open in browser:
- Web UI: http://127.0.0.1:8000
- API Docs: http://127.0.0.1:8000/docs
# Build image
docker build -t ats-resume-app .
# Run container
docker run -p 8000:8000 -e OPENAI_API_KEY=sk-your-key ats-resume-appThis app can be deployed to various platforms:
- Railway (Recommended) - Best for FastAPI apps
- Render - Easy deployment with auto-scaling
- Vercel - Serverless functions (limited for long-running tasks)
- Fly.io - Global distribution
Quick Deploy to Railway:
- Push code to GitHub
- Go to https://railway.app and sign up with GitHub
- Create new project → Deploy from GitHub repo
- Add
OPENAI_API_KEYenvironment variable in Variables tab - Railway auto-deploys! Get your URL from Settings → Domains
Detailed Railway Setup:
- Connect your GitHub repository to Railway
- Railway automatically detects FastAPI and deploys
- Add environment variables in the Variables tab
- Your app will be live at
https://your-app.up.railway.app
GET /- Main HTML interface
POST /generate_resume/- Generate resume from form data (returns DOCX file)POST /api/generate_resume- Generate resume from JSON (returns metadata + download link)GET /download/{filename}- Download generated resumeGET /health- Health check endpointGET /docs- Interactive API documentation (Swagger UI)
# JSON API
curl -X POST "http://localhost:8000/api/generate_resume" \
-H "Content-Type: application/json" \
-d '{
"job_description": "Looking for a Python developer with FastAPI experience..."
}'- Basic FastAPI server
- LLM keyword extraction (placeholder)
- DOCX resume generation
- Simple web UI
- Docker support
- OpenAI Integration - Real LLM keyword extraction
- Enhanced Resume Generation - Better formatting, STAR bullets
- Vector Search (RAG) - Match experience with job requirements
- PostgreSQL Database - User accounts, job history
- JWT Authentication - Secure multi-user access
- Logging & Metrics - Structured logging, Prometheus metrics
- Multiple Resume Templates - Choose from C3, modern, creative
- Batch Processing - Generate resumes for multiple job descriptions
- Resume Analytics - ATS score, keyword match percentage
- CI/CD Pipeline - Automated testing and deployment
- Cloud Deployment - Deploy to AWS/GCP/Azure
# Run tests (when implemented)
pytest tests/
# With coverage
pytest --cov=src tests/| Variable | Description | Required |
|---|---|---|
OPENAI_API_KEY |
OpenAI API key for LLM features | Yes (for production) |
GEMINI_API_KEY |
Google Gemini API key (alternative) | Optional |
DATABASE_URL |
PostgreSQL connection string (future) | No |
JWT_SECRET |
Secret key for JWT auth (future) | No |
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
This project is licensed under the MIT License.
- Built with FastAPI
- LLM powered by OpenAI / Google Gemini
- Document generation using python-docx
For questions, issues, or suggestions, please open an issue on GitHub.
Built with ❤️ for Full-Stack and ML/LLM Engineers