MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202611090809 A) filed by Noida Institute Of Engineering And Technology Niet on July 26, 2026, for Privacy-Preserving Federated Computing System And Method For Employee-Aware Enterprise Digital Transformation Pacing With Motivation-Risk Analytics.
Inventors include Srishti; Shruti Srivastava; and Dr. Shagun Sharma.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: The present invention discloses a privacy-preserving federated computing system (100) and method (300) comprising edge devices (105) with neural processing units and a central server (115) with tensor processing units coupled through a network (200). Each edge device (105) executes a data acquisition module (110), an employee profile engine (120), and a local motivation-risk analytics module (150) computing on-device motivation-risk indices via a calibrated logistic function combined with isolation forest and long short-term memory autoencoder anomaly detection. The central server (115) executes a justice perception modeling module (130), a motivation state inference engine (140) with a means-ends fusion scorer (144), and a transformation pacing recommender (160) implementing a LinUCB contextual bandit. A feedback loop (170) performs federated retraining via secure aggregation of differentially private updates. Technical effects include reduced network telemetry, lower compute load, sub-500-millisecond alerts, and cryptographic privacy.
Disclaimer: Curated by HT Syndication.
