Skip to content

Siddartha-Kodaboina/HR-Resume-Screener

Repository files navigation

🎯 LLM-Powered Resume Matching System

📝 Problem & Solution

Transformed traditional resume screening by developing an AI-driven solution that reduces candidate screening time by 5 hours. The system enables recruiters to perform natural language queries for candidate matching, replacing manual filtering with intelligent, automated processing.

🏗️ Architecture

  • Resume Parsing: LlamaParse for structured metadata extraction
  • Vector Database: Dual embedding strategy in Pinecone with static chunking (200 tokens, 15-char overlap)
  • Language Models: LLaMA 3.2 & GPT-4 for semantic analysis
  • Retrieval: RAG-based system with Ada Embeddings
  • Frontend: Streamlit UI for intuitive interaction

🛠️ Tools & Technologies

  • LlamaIndex & LlamaCloud: for orchestration
  • Pinecone: for vector storage
  • GPT-4: for candidate matching
  • LlamaParse: for document processing
  • Streamlit: for user interface
  • Python: for backend development

✨ Key Achievement

Reduced recruitment screening time by 5 hours through intelligent automation and semantic search capabilities.

🚀 Getting Started

1. API Key Configuration

Setup your open AI key in .env file

2. Starting the Django Server

python manage.py runserver 9000

3. Starting the chainlit App

In another terminal, start the chainlit app

chainlit run app.py -w

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published