Enhance Recruiters Productivity and Search Quality Using Machine Learning
October 27, 2020
American multinational human resource consulting firm
Problem Statement
- Legacy systems need to be replaced to aid faster delivery cycles, meeting higher %age of SLAs, and greater insights into candidate profiles.
- To develop sophisticated IT and Data Science infrastructure to enable end-to-end talent acquisition for clients right from receiving job orders to finding candidates
Solution Approach
- Candidate Halo ingests data from Oracle data base
Data is stored in XML object format
Identify clusters within 800 job titles
Unsupervised machine learning models
Integrated Machine Learning Model
NLP framework to identify key concepts from conceptual job description
Business Impact
End-to-End SLA from search to match is 6-10 seconds for complex search
Improvement in Employability rate from 1.5% to 3.0% post production
Expanded candidate identification, richer job/skills mapping
Critical Success Factors
- Feature Extraction
- Candidate Halo scores candidates In near-real-time
