Intellectual Property India Publishes Patent Application for Machine Learning-Based Cloud Platform For Real-Time Fraud Detection In Distributed Banking Networks

MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641078411 A) filed by Chintolla Surekha; Dr M Ramakrishnan; Dr. Manohara H T; F. Josephine Lenta; M Sameena Nazeer; Dondeti Sowmya; Mattaparthi Sailaja; Dr. D. Sowjanya; Dr Sumanta Bhattacharya; P. Veeramani; Akarsha D P; and Hemalbhai Kanubhai Patel on June 25, 2026, for Machine Learning-Based Cloud Platform For Real-Time Fraud Detection In Distributed Banking Networks.

Inventors include Chintolla Surekha; Dr M Ramakrishnan; Dr. Manohara H T; F. Josephine Lenta; M Sameena Nazeer; Dondeti Sowmya; Mattaparthi Sailaja; Dr. D. Sowjanya; Dr Sumanta Bhattacharya; P. Veeramani; Akarsha D P; and Hemalbhai Kanubhai Patel.

The application for the patent was published on July 31, 2026, under issue no. 31/2026.

Abstract: Machine Learning-Based Cloud Platform for Real-Time Fraud Detection in Distributed Banking Networks is the proposed invention. The proposed invention is designed to identify and prevent fraudulent financial activities by continuously monitoring transaction data generated across multiple banking channels, including internet banking, mobile banking, ATM networks, payment gateways, and card processing systems. The proposed platform employs a hybrid machine learning architecture comprising a Variational Autoencoder (VAE) and a Long Short-Term Memory (LSTM) network to analyze transaction behaviours and detect anomalies in real time. The VAE learns normal transaction patterns and identifies suspicious activities through anomaly detection, while the LSTM network captures temporal dependencies and customer behavioural trends to predict potential fraud. A cloud-based analytics engine performs data preprocessing, feature extraction, behavioural profiling, and dynamic fraud risk assessment on large-scale transaction streams. Based on calculated fraud risk scores, the system automatically initiates appropriate responses, including transaction approval, enhanced authentication, manual review, or transaction blocking. The invention provides improved fraud detection accuracy, reduced false-positive alerts, enhanced customer security, minimized financial losses, and intelligent real-time protection for modern digital banking ecosystems.

Disclaimer: Curated by HT Syndication.