MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202641123056 A) filed by Vellore Institute Of Technology on October 14, 2026, for Database-Native Machine Learning Experiment Tracking System With Trigger-Based Deduplication And Recursive Lineage.
Inventors include Sangeetha S; Suyash Chetan Dabholkar; Yashashwi Jain; and Rakshit Sinha.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: The present disclosure proposes a database-native machine-learning experiment tracking system (100) for recording, monitoring, comparing, ranking, recovering, tracing, and querying model-training experiments and associated experiment data. The system (100) comprises a computing device (102) having a processor (104) and a memory (106) storing instructions executable by the processor (104). The processor (104) executes a plurality of modules (108) comprising a parameter fingerprint module (110), a duplicate-run detection module (112), an anomaly detection module (114), a composite scoring module (116), a failed-run preservation and revival module (118), a lineage reconstruction module (120), and an experiment-history retrieval module (122). The system (100) evaluates completed model-training runs using performance parameters and configurable metric weights and enables controlled retrieval of stored experiment information while restricting unintended modification of the stored experiment information.
Disclaimer: Curated by HT Syndication.
