MUMBAI, India, Jan. 23 -- Intellectual Property India has published a patent application (202511126826 A) filed by Shahrukh Khan; Rohan Singh Bais; Dr. Pradeep Sharma; Gokul Ram Nishad; Prof. Md. Sadique Shaikh; Dr. Jay Upendra Dhruv; Dr. Hardik Jayendrakumar Prajapati; Dr. Shradhdha Vinodbhai Thakkar; and Yogesh Kumar, Ghaziabad, Uttar Pradesh, on Dec. 15, 2025, for 'a system and method for ai-enabled software optimization and code performance analysis.'
Inventor(s) include Shahrukh Khan; Rohan Singh Bais; Dr. Pradeep Sharma; Gokul Ram Nishad; Prof. Md. Sadique Shaikh; Dr. Jay Upendra Dhruv; Dr. Hardik Jayendrakumar Prajapati; Dr. Shradhdha Vinodbhai Thakkar; and Yogesh Kumar.
The application for the patent was published on Jan. 23, under issue no. 04/2026.
According to the abstract released by the Intellectual Property India: "The disclosure describes a system and method that utilize AI in a software and hardware optimization context and in performance analysis of code for specific hardware. The system includes a code ingestion and semantic extraction module that produces abstract syntax trees (or trees representative of code structure), a hardware-anchored semantic execution graph engine that associates semantic data with processor telemetry, and a dynamic instruction-level self-amending optimizer that uses processor telemetry to create runtime patches at the micro-architectural level and execute changes that optimize thread-level code performance. A quantum-inspired code state reducer speeds up the generation of optimizations by reducing high-dimensional execution traces, while a latency-aware system-call governor predicts and reduces I/O delays. The method accepts code, collects hardware metrics, generates the execution graph, identifies bottlenecks, generates and reviews for accuracy optimization patches in a "sandbox," and utimately deploys authorized changes into the execution environment also maintaining a continuous learning repository of optimization signatures for one platform that can be retrained for use on another platform. The disclosure provides a real-time adaptive method for enhancing performance that is specific to hardware system and represents an improvement over prior art compilers and profilers."
Disclaimer: Curated by HT Syndication.
