Deepseek Enabled Job Search Tool With Vectorized DB
A semantic search engine for job postings with LLM…
In this project, I utilized the JobSpy open-source tool to scrape job listings that match specific search criteria. These listings are stored in a MariaDB database. I then employed SentenceTransformers to generate FAISS embeddings, enabling fast nearest neighbors searches. By matching user queries against both a vectorized version of my resume and potential job listings, I leverage deepseek to analyze and respond to these queries. This setup allows me to explore the types of jobs available and assess my qualifications for them.
Screenshot 1: Overview of the semantic search engine interface.
Screenshot 2: Detailed view of job matching results.
Desired Improvements
- Enhance the search algorithm to improve accuracy and relevance of job matches.
- Integrate user feedback to refine and optimize the matching process.