Patents by Inventor Divya Dhar

Divya Dhar has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Method and system for evaluating effectiveness of prompt
    Patent number: 12681969
    Abstract: Method, system, and computer-readable storage media for evaluating effectiveness of a prompt are disclosed. Input data for generating the prompt is received. Based on the input data, sub-prompts are generated and each of the sub-prompts is evaluated. For evaluating a sub-prompt, a sub-response corresponding to the sub-prompt is generated, and an optimal score is computed for the sub-response. The optimal score value is computed for the sub-response by computing a score value for evaluation of each evaluation aspect of multiple evaluation aspects and computing a combined score value based on the score value computed for evaluation of each evaluation aspect. The optimal score value is compared with a predetermined threshold value. If the optimal score value exceeds the predetermined threshold value, the respective sub-prompt is determined as the most effective prompt for generating a response using a Large Language Model (LLM) corresponding to the input data.
    Type: Grant
    Filed: December 13, 2024
    Date of Patent: July 14, 2026
    Assignee: Accenture Global Solutions Limited
    Inventors: Aditya Kumar, Hemant Chandrakant Patil, Sarang Padmakar Joshi, Ninad Shashikant Kulkarni, Niladri Sekhar Dutta, Karthik Ramesh, Sailaja Bandi, Divya Dhar, Srenik Ravjiani
  • METHOD AND SYSTEM FOR EVALUATING EFFECTIVENESS OF PROMPT
    Publication number: 20260170029
    Abstract: Method, system, and computer-readable storage media for evaluating effectiveness of a prompt are disclosed. Input data for generating the prompt is received. Based on the input data, sub-prompts are generated and each of the sub-prompts is evaluated. For evaluating a sub-prompt, a sub-response corresponding to the sub-prompt is generated, and an optimal score is computed for the sub-response. The optimal score value is computed for the sub-response by computing a score value for evaluation of each evaluation aspect of multiple evaluation aspects and computing a combined score value based on the score value computed for evaluation of each evaluation aspect. The optimal score value is compared with a predetermined threshold value. If the optimal score value exceeds the predetermined threshold value, the respective sub-prompt is determined as the most effective prompt for generating a response using a Large Language Model (LLM) corresponding to the input data.
    Type: Application
    Filed: December 13, 2024
    Publication date: June 18, 2026
    Applicant: Accenture Global Solutions Limited
    Inventors: Aditya KUMAR, Hemant Chandrakant PATIL, Sarang Padmakar JOSHI, Ninad Shashikant KULKARNI, Niladri Sekhar DUTTA, Karthik RAMESH, Sailaja BANDI, Divya Dhar, Srenik Ravjiani