💼 AI Job Recommender System (Real-Time + Semantic Matching)

An intelligent job recommendation system that scrapes live job postings and delivers personalized job matches using semantic similarity powered by transformer models.

---

🚀 Overview

This project helps job seekers discover relevant opportunities by combining:

• 📡 Real-time job scraping
• 📄 Resume analysis
• 🧠 Transformer-based NLP matching

Unlike traditional systems that rely on keyword matching, this system understands context and meaning, providing more accurate recommendations.

---

✨ Key Features

🔍 Real-Time Job Scraping

• Dynamically fetches job listings from RemoteOK, Adzuna Api
• Ensures up-to-date job recommendations

📄 Resume Parsing

• Extracts skills and relevant information from uploaded resumes
• Supports structured and unstructured formats

🧠 Semantic Recommendation Engine

• Uses Sentence Transformers to generate embeddings
• Computes Cosine Similarity to rank job relevance

🎯 Flexible Input System

• Upload resume OR input skills manually
• Combines both for improved accuracy

⚡ On-Demand Processing

• Recommendations are generated only when triggered
• Efficient and scalable design

---

🏗️ System Workflow

User Input / Resume
        ↓
Resume Parsing (NLP)
        ↓
Skill Extraction
        ↓
Job Scraper (Live Data)
        ↓
Text Embedding (Sentence Transformers)
        ↓
Similarity Matching (Cosine Similarity)
        ↓
Top Job Recommendations

---

🛠️ Tech Stack

• Python 3.11
• Streamlit – Interactive UI
• Pandas / NumPy – Data processing
• Sentence Transformers – Semantic embeddings
• Scikit-learn – Cosine similarity
• Requests – Web scraping

---

📂 Project Structure

Job_Recommender/
│
├── app.py                    # Streamlit app entry point
├── job_scraper.py            # Job scraping logic
├── recommender.py            # Recommendation engine
├── requirements.txt
└── README.md

---

⚙️ Installation

# Clone repository
git clone https://github.com/Codaks-py/Job_Recommender.git

# Navigate into project
cd Job_Recommender

# Create virtual environment
python -m venv venv

# Activate environment
venv\Scripts\activate      # Windows
source venv/bin/activate   # Mac/Linux

# Install dependencies
pip install -r requirements.txt

---

▶️ Usage

streamlit run app.py

Steps:

1. Upload your resume OR enter skills manually
2. Click "Get Recommendations"
3. View ranked job matches instantly

link: https://jobrecommenderapp.streamlit.app/

---

📊 Output

The system returns:

• 💼 Job Title
• 🏢 Company
• 📈 Match Score (%)
• 🧠 Relevant Skills
• 🔗 Job Link

---

🧠 How the Recommendation Works

This system uses semantic similarity powered by transformer models:

• Text from resumes and job descriptions is encoded using Sentence Transformers
• Each input is converted into a dense vector (embedding)
• Cosine similarity is used to measure closeness
• Jobs are ranked based on semantic relevance

✅ Why This Approach?

• Understands context, not just keywords

• Matches similar meanings:

  • "ML Engineer" ≈ "Machine Learning Developer"

• Significantly more accurate than traditional methods

---

🧠 Key ML Highlight

This project demonstrates a real-world NLP application using:

• Transformer-based embeddings
• Semantic search / retrieval systems
• End-to-end ML pipeline integration

Similar techniques are used in modern platforms like LinkedIn and Indeed.

---

🔥 Why This Project Stands Out

• Combines ML + NLP + Data Engineering + UI
• Uses real-time job data (not static datasets)
• Implements transformer-based recommendation system
• Built as an end-to-end deployable application

---

🚀 Future Improvements

• 🌐 Multi-source job scraping
• 🤖 Upgrade to advanced embeddings (fine-tuned models)

---

🤝 Contributing

Contributions are welcome!

git checkout -b feature-name
git commit -m "Add feature"
git push origin feature-name

---