---
title: ATS Resume Optimizer
emoji: 📄
colorFrom: red
colorTo: indigo
sdk: streamlit
sdk_version: 1.58.0
app_file: src/streamlit_app.py
pinned: false
---



# 📄 ATS Resume Optimizer

An AI-powered tool that analyzes your resume against any job description and tells you how well it matches — plus what keywords you're missing. Built with **Streamlit**, **MiniLM embeddings**, and **spaCy NLP**.

---

## 🚀 Try It Now

Paste your resume and a job description. Get an instant **ATS match score** and **keyword suggestions**.

No signup required. Runs entirely in your browser.

---

## ✨ What It Does

| Feature | How It Helps |
|---------|-------------|
| 🧠 **Semantic Matching** | Understands meaning, not just keyword counting. "Led team" and "managed group" score the same. |
| 🔍 **Keyword Gap Analysis** | Extracts skills, tools, and responsibilities from the job description and shows what's missing from your resume. |
| 📊 **ATS Score (0-100%)** | Clear match percentage with color-coded feedback: 🟢 High (>75%), 🟡 Medium (50-75%), 🔴 Low (<50%). |
| 📝 **5 Demo Profiles** | Test instantly with pre-loaded resumes for Data Scientist, Frontend Dev, PM, Marketing, and ML Engineer. |

---

## 🛠️ How It Works

```
Resume + Job Description
         ↓
   [MiniLM Transformer]
         ↓
   Embedding Vectors
         ↓
   Cosine Similarity → ATS Score
         ↓
   [spaCy NLP] → Missing Keywords
         ↓
   Feedback + Suggestions
```

**Model:** [`sentence-transformers/all-MiniLM-L6-v2`](https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2) — 22MB, fast on CPU, trained on 1B+ sentence pairs.

---

## 📊 Example Output

**Input:** Senior Data Scientist resume vs. matching job description

**Result:**
- 🟢 **ATS Match Score: 87%** — HIGH MATCH. Ready to submit!
- 📝 **Missing Keywords:** `stakeholder`, `a/b testing`, `kubernetes` *(consider adding these)*

---

## 🎮 Quick Start

1. **Select a demo** from the sidebar (or paste your own text)
2. **Paste your resume** in the left box
3. **Paste the job description** in the right box
4. **Click "Analyze"** and get instant feedback

---

## 🏗️ Tech Stack

| Component | Technology |
|-----------|-----------|
| **UI** | Streamlit |
| **Embeddings** | MiniLM-L6-v2 (Hugging Face) |
| **NLP** | spaCy `en_core_web_sm` |
| **Backend** | PyTorch |

---

## 📁 Repo Structure

```
├── src/
│   └── streamlit_app.py      # Main app
├── Dockerfile                # Container config
├── requirements.txt          # Dependencies
└── README.md                 # This file
```

---

## 🌐 Deployment

This Space runs on Hugging Face's free tier. MiniLM auto-downloads on first launch (~22MB). No GPU needed.

**To run locally:**
```bash
git clone https://huggingface.co/spaces/yourusername/ats-resume-optimizer
cd ats-resume-optimizer
pip install -r requirements.txt
python -m spacy download en_core_web_sm
streamlit run src/streamlit_app.py
```

---

## 📝 Notes

- **Privacy:** All text processing happens locally in the Space. No data is stored or sent to external APIs.
- **First Load:** May take 10-20 seconds on first visit while MiniLM downloads.
- **spaCy Model:** Downloaded at build time via Dockerfile.

---

## 🔮 Roadmap

- [ ] PDF resume upload
- [ ] Section-by-section comparison (Skills vs Skills, Experience vs Requirements)
- [ ] AI-powered rewrite suggestions (GPT integration)
- [ ] Chrome extension for LinkedIn/Indeed job descriptions

---

## 👤 Author

**Muhammad Ibtihaj** — ML Engineer & Full-Stack Developer

---

<p align="center">
  <b>Made with ❤️ and Transformers 🤖</b><br>
  <i>Land your dream job, one keyword at a time! 📄✨</i>
</p>