MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202611109873 A) filed by Ramesh Bishukarma on September 13, 2026, for Agentic Artificial Intelligence System And Method For Autonomous Cloud Cost-Anomaly Detection And Financial Operations Optimization.
Inventor includes Ramesh Bishukarma.
The application for the patent was published on September 18, 2026, under issue no. 38/2026.
Abstract: An agentic artificial intelligence system and computer-implemented method are disclosed for autonomous cloud cost-anomaly detection and financial operations optimization. Heterogeneous billing, resource, pricing, deployment, ownership and operational telemetry are normalized into a canonical cost-resource representation and associated through a context graph. Multiple anomaly detectors establish context-aware dynamic baselines and produce a composite anomaly confidence. Cooperating autonomous agents generate evidence-linked causal hypotheses and candidate optimization actions. A policy and risk governance engine determines an autonomy class using confidence, reversibility, resource criticality, blast radius and operational constraints, thereby controlling recommendation, approval or automatic execution. Authorized actions are optionally staged and executed through cloud interfaces. A verification and rollback engine measures workload-normalized realized savings and operational effects, initiates rollback when a guardrail is violated, and feeds verified outcomes back to recalibrate subsequent anomaly and optimization decisions. A tamper-evident decision record preserves evidence, governance and execution history. Accompanied Drawing [FIGS. 1-2].
Disclaimer: Curated by HT Syndication.
