MUMBAI, India, Sept. 7 -- Intellectual Property India has published a patent application (202641098895 A) filed by Kgr Foundation For Business Incubation on August 15, 2026, for A Machine Learning Architecture For Attrition Risk Stratification.
Inventors include Chandana Chilaka; Dr. K. Sarika; and P Manikanta Babu.
The application for the patent was published on September 04, 2026, under issue no. 36/2026.
Abstract: A MACHINE LEARNING ARCHITECTURE FOR ATTRITION RISK STRATIFICATION A novel two-stage machine learning architecture for predicting employee attrition risk prior to a candidate's first day of employment. Unlike existing human resources analytics that rely exclusively on post-hire data, the disclosed method captures pre-employment negotiation session dynamics. A Negotiation Information Extraction module encodes behavioral signals into specific mathematical features (ORI, CFS, UMC, NMDI). These are processed by a Negotiation Behavior Estimator to isolate negotiation risk, which an Attrition Risk Classifier then fuses with standard HR data. This pipeline generates a pre-hire Attrition Risk Score at the moment a negotiation closes. The system enables proactive HR interventions, targeted onboarding, and continuous longitudinal risk monitoring, providing a critical new window for attrition management.
Disclaimer: Curated by HT Syndication.
