[Resume Intelligence, Match Scoring]
Smart Resume Matcher
“Smart Resume Analyzer & Job Matcher” parses resumes and instantly matches them to job postings using TF-IDF, Word2Vec, and Sentence-BERT embeddings with FAISS for sub-second vector search.
Overview
The system ingests PDF/DOCX resumes, extracts skills/experience/education, encodes text with multiple representations, and performs cosine similarity with FAISS across indexed job descriptions. Weighted scoring powers clear, real-time match insights for candidates and recruiters.
Key Features
- Resume parsing with skills/experience/education extraction
- TF-IDF, Word2Vec, and Sentence-BERT embeddings for matching
- FAISS-powered sub-second vector search with weighted scoring
- Clear match insights for candidates and recruiters
Technologies Used
- Backend: Python 3.9+ with Flask/FastAPI
- NLP Pipeline: spaCy, NLTK, Sentence-Transformers (all-mpnet-base-v2)
- Vector Search: FAISS for similarity queries
- Database: MongoDB for parsed data & embeddings
- Frontend: React.js + Material-UI dashboard
- Text Extraction: PyPDF2 & python-docx for PDF/DOCX parsing
Contributors
Ameya Kolhatkar • Andrew Loh • Faith Han • Senuvi Jayasinghe
Demo Video
Project Stills
