🧠 AI Resume Screening Agent

This project is an AI-powered Resume Screening Agent that ranks candidate resumes based on how well they match a given Job Description (JD).

Unlike many AI agents, this one is:

• ✅ Completely offline (no external APIs)
• ✅ 100% free to run (no OpenAI / Claude / Gemini keys)
• ✅ Fast and lightweight
• ✅ Built with Streamlit + scikit-learn

---

🚀 What It Does

1. You paste a Job Description.
2. You upload one or more PDF resumes.
3. The agent:
   • Extracts text from each resume.
   • Uses TF-IDF + Cosine Similarity to measure how close each resume is to the JD.
   • Calculates a score (0–100) for each resume.
   • Identifies basic strengths (matching keywords) and gaps (missing keywords).
4. It then shows a ranked list of candidates from best match → lowest match.

This is ideal as a first-pass resume filter in an HR process.

---

🧱 Tech Stack

• Language: Python
• UI: Streamlit
• ML / Scoring:
  • scikit-learn (TF-IDF, cosine similarity)
• PDF Text Extraction:
  • PyPDF2
• APIs Used: None (fully offline ML, no external services)

---

🏗 Project Structure

resume-screening-agent/
├── app.py           # Streamlit UI
├── screening.py     # Core logic: PDF reading + TF-IDF scoring
├── requirements.txt # Python dependencies
└── venv/            # Local virtual environment (not needed on GitHub ideally)