LinkedIn Job Matcher
Your personal AI-powered job hunting assistant that finds perfect job matches while you sleep.
Project Details
- Client: Self-initiated
- Year: 2024
- Duration: 3 weeks
What We Built
This tool automates the job search process by intelligently scanning LinkedIn, matching opportunities to your CV, and delivering a curated list of relevant jobs. The system operates in a continuous loop to ensure you never miss a new opportunity. It starts by parsing your CV to understand your skills and experience, then performs an initial scan of LinkedIn for recent job postings based on your preferences. After the first scan, it saves a checkpoint to avoid re-analyzing the same jobs. Each new job is analyzed by a rotating set of AI models to determine a match score, acting as a hiring manager comparing the job description against your CV. The system then monitors for new jobs every 10-15 minutes, and jobs with a match score above 70% are saved and a notification is sent.
The Problem
The traditional job search is a time-consuming and inefficient chore. Manually sifting through hundreds of job postings on platforms like LinkedIn leads to missed opportunities and hours of wasted effort. The challenge was to create a system that could automate this entire process, from discovery to qualification, ensuring that users only see the most relevant job openings without lifting a finger.
How We Solved It
The solution is a modular Python application designed for clarity and extensibility. It uses Playwright for robust browser-based scraping of LinkedIn, PyPDF2 to extract text from the user's CV, and the OpenAI library to connect to OpenRouter for AI-powered matching. This setup rotates through various free, high-performance language models, ensuring resilience and avoiding dependency on a single provider. The core logic is orchestrated by main.py, with separate modules for configuration, scraping, CV parsing, AI matching, storage, and notifications. This clean architecture makes the system easy to maintain and extend.
Built With
Screenshots
