Resume2Job

Resume2Job is a learning-focused full-stack project that analyzes how well a resume matches a given job description.

It helps users clearly understand their strengths, gaps, and improvement areas using structured AI-generated feedback without vague or essay-style responses.

This project is built to demonstrate real-world LLM integration, document processing, and frontend–backend coordination.
It is not a commercial product, but a solid engineering + portfolio project.

---

🚀 Features

• Upload Resume (PDF / DOCX / Image)
• Upload or paste Job Description
• Automatic resume & JD text extraction
• Cleaning, sectioning, and chunking of content
• Context-aware AI analysis
• Structured, readable output (no essay dumping)
• Adaptive response depth:
  • Short answers by default
  • Detailed guidance only when explicitly asked
• Fully working frontend + backend
• Free-tier friendly (no paid APIs required)

---

📸 Screenshots

📱 Mobile View

🧠 How the System Works

1. User uploads a resume and a job description
2. Backend extracts raw text from uploaded files
3. Text is cleaned and split into logical sections
4. Important content chunks are stored per session
5. User asks a question (e.g. "Am I qualified for this role?")
6. Relevant resume & JD chunks are injected into an AI prompt
7. Gemini generates a structured, career-focused response
8. Frontend renders the output using Markdown

---

🏗️ Tech Stack

Frontend

• React
• Fetch API
• React Markdown (clean formatted output)

Backend

• Python (Flask)
• Flask-CORS
• Background threading for processing
• Session-based in-memory storage

AI / LLM

• Google Gemini (gemini-pro)
• Prompt engineering for output control
• Embeddings intentionally disabled (free-tier safe)

---

📁 Project Structure

resume2job/
│
├── backend/
│   ├── app.py                # Flask API entry point
│   ├── extraction/           # Resume text extraction logic
│   ├── processing/           # Cleaning, sectioning, chunking
│   ├── matching/             # JD parsing & in-memory storage
│   ├── utils/                # File handling helpers
│   └── requirements.txt
│
├── frontend/
│   ├── src/
│   │   └── Resume2Job.jsx    # Main UI logic
│   └── package.json
│
└── README.md

✨Response Intelligence (Important Design)

The system controls response length and depth, not the model.

• Default Behavior
• Short, scannable responses
• Clear verdict, strengths, and gaps

When the User Asks for Details (e.g. "What should I do to guarantee qualification?")

• Deeper analysis
• Actionable improvement steps
• Structured recommendations
• This is handled using prompt-level response modes, not by switching models.

🎯Project Goals

• This project was built to:
• Learn end-to-end LLM integration
• Practice backend architecture & debugging
• Understand prompt engineering deeply
• Build a deployable AI system
• Focus on clarity, correctness, and stability
• This is a learning + portfolio project, not a commercial product.

⚠️Limitations

• No semantic similarity search (embeddings disabled)
• Session data is not persistent
• Output quality depends on input quality
• Not optimized for high concurrent users
• These trade-offs were made intentionally to keep the system simple, reliable, and free-tier friendly.

👤 Author

Danish Shaikh BCA Student | Python & Data-Oriented Developer

📄 License

This project is intended for educational and personal learning purposes.