MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202521039268 A) filed by Lti Mindtree Ltd. on April 23, 2025, for Method And System For Optimizing Prompts For Large Language Models (llms).
Inventors include Sahithi Priya Vadlakonda; T Siva Rama; Sateesh Gottumukkala; and Ashish Varerkar.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: A method and system (102) for optimizing prompts for LLMs is disclosed. The system (102) receives a baseline prompt (108) configured for interacting with an LLM (112), and transforms it into a structured prompt in accordance with a standardized architecture by utilizing prompt engineering templates. An iterative prompt optimization process is performed that comprises evaluating performance of the structured prompt against a ground truth dataset comprising query-response pairs. A plurality of prompt variations for the structured prompt are generated based on a performance metrics obtained from the evaluation. The system (102) refines structured prompt by incorporating elements from the one or more optimal prompt variations. The refined structured prompt is stored in a centralized prompt management system (104) that maintains version control for a plurality of prompts, and a user interface (106) is provided that enables users to interact with the LLM (112) using the refined structured prompt.
Disclaimer: Curated by HT Syndication.
