swarajaya/user-behavior-analytics

🚀 AI Resume Screening & Ranking System (SaaS)

An End-to-End AI SaaS application that screens and ranks resumes based on a job description using ML + NLP + LLM. Built with Python, deployed with Streamlit, and powered by AI for candidate matching.

🔍 Features

• —Upload multiple resumes (PDF)
• —Upload job description (PDF)
• —Automatic skill extraction from resumes
• —Compare candidate skills with job description
• —Assign match score (0–1)
• —Rank candidates automatically
• —Optional: AI explanation for candidate suitability (LLM)
• —Optional: Download ranked results

🛠 Tech Stack

• —Python – Main language
• —Streamlit – Frontend SaaS app
• —PyPDF2 – PDF text extraction
• —Pandas / NumPy / scikit-learn – ML & data processing
• —spaCy – NLP (optional advanced skill extraction)
• —OpenAI GPT – Candidate reasoning (optional)
• —Hugging Face Spaces – Deployment

📁 Folder Structure

ai-resume-screening/ │ ├── app.py # Streamlit app ├── requirements.txt # Dependencies ├── README.md # Project documentation ├── src/ # Python modules │ ├── resumeparser.py │ ├── jdparser.py │ ├── skillextractor.py │ ├── matcher.py │ ├── ranker.py │ └── llmhelper.py ├── data/ # Sample demo files (optional) │ ├── resumes/ │ │ └── sampleresume.pdf │ └── jobdescriptions/ │ └── sample_jd.pdf └── .gitignore

⚡ How It Works

1. 1.Extract text from uploaded PDFs
2. 2.Extract skills from resumes & job description
3. 3.Compare skills and calculate match score
4. 4.Rank candidates by score
5. 5.Optional: Generate AI explanation for each candidate

💻 Usage

1. 1.Clone the repo:

git clone https://github.com/<your-username>/ai-resume-screening.git
cd ai-resume-screening ai-resume-screening/
│
├─ app.py # Streamlit app
├─ ranker.py # Ranking & matched skills logic
├─ llm_helper.py # Optional AI explanation helper
├─ requirements.txt # Dependencies
├─ uploads/ # Folder for uploaded resumes (auto-created)
└─ README.md # Project description & instructions


---

## **💡 Why This Project is Useful**
- **Saves time** for recruiters by automating candidate shortlisting.  
- Provides **insights into matched skills**, making hiring decisions data-driven.  
- Demonstrates **Python, Streamlit, and NLP text-processing skills**.  
- Fully deployable and **production-ready** on Hugging Face Spaces.  

---

## **📌 Optional Enhancements**
- Integrate **OpenAI or Hugging Face LLM** for candidate evaluation text.  
- Add **visual charts** for top candidates and skill coverage.  
- Include **predefined skill lists** for more accurate ranking.  

---

*Try it now! Upload your JD and resumes, and see the AI-powered ranking in action.*