Skip to content

Resume–JD Alignment Feature: Enhanced Skills Extraction, Scoring, Skills matching accuracy - #7

Open
ashutoshsom1 wants to merge 5 commits into
onepointconsulting:mainfrom
ashutoshsom1:main
Open

Resume–JD Alignment Feature: Enhanced Skills Extraction, Scoring, Skills matching accuracy#7
ashutoshsom1 wants to merge 5 commits into
onepointconsulting:mainfrom
ashutoshsom1:main

Conversation

@ashutoshsom1

Copy link
Copy Markdown

Enhanced resume–JD alignment logic for better skills and education matching. Improves candidate ranking and user experience.

Changes

  • JD → CV alignment logic
  • Updated LLM prompts
  • Data model updates
  • UI improvements (ranking view)
  • Added/updated tests
  • Refactoring

Testing

  • Manual: Upload JD & CV, verify ranking and breakdown.
  • Automated: Updated skill/education extraction tests.

Checklist

  • Builds locally
  • Tests pass
  • No secrets committed
  • Docs updated

This commit includes:
- Fixed Azure OpenAI integration by updating config.py to use AzureChatOpenAI with proper deployment parameters
- Enhanced PDF processing with proper poppler path configuration in pdf2image_converter.py
- Improved error handling in matcher_chainlit.py for more robust processing of skills and education data
- Refactored process_generic_extraction function to handle different types of extraction with consistent error handling
- Added proper environment configuration in code.env with Azure OpenAI credentials and system settings
- Implemented complete end-to-end workflow for uploading, processing, and matching job descriptions with candidate CVs
- Added visualization capabilities for candidate rankings with detailed breakdowns of skills and education matches

The system now successfully processes multiple job descriptions and CVs, extracts relevant skills and education information, and provides a comprehensive ranking of candidates based on configurable scoring weights.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant