Job Scraper and Analyzer

A Python-based web scraping and data analysis tool designed to archive and analyze job advertisements from major Central European job boards. This project leverages Local LLMs to extract insights from job descriptions without compromising data privacy.

Features

• Supported job portals:
  • Hungary: profession.hu, cvonline.hu
  • Slovakia: profesia.sk
• Data Archiving: Automatically saves every job description into a raw .txt file for permanent backup, ensuring you have a record even after the ad expires.
• AI-Powered Analysis: Uses the Mistral-Nemo model running locally via Ollama to parse requirements, tech stacks, and benefits from unstructured text.
• Structured Output: Generates a comprehensive CSV database containing metadata (salary, location, company).
• Privacy: No job data is sent to external APIs; all processing happens on your local machine.

---

Tools Used

• Language: Python 3.13.x
• Libraries: BeautifulSoup4, Requests, Pandas
• Local AI: Ollama
• Model: mistral-nemo (12B parameters, 8-9GB VRAM recommended)

---

Project Structure

├── CSV/                 # Scraped data in .csv file (profession_jobs.csv)
├── Raw text/            # Permanent backup of job descriptions (.txt)
└── scraper.py           # Scraping scripts

How to run