MUMBAI, India, Aug. 24 -- Intellectual Property India has published a patent application (202621080937 A) filed by Symbiosis International Deemed University on June 30, 2026, for Autonomous Multi-Agent Retrieval-Augmented Generation System For Scientific Document Understanding.
Inventors include Jay Thakkar; Varshil Prajapati; Dr. Aniket K. Shahade; Dr. Priyanka V. Deshmukh; Dr. Shruti Patil; and Dr. Ketan Kotecha.
The application for the patent was published on August 21, 2026, under issue no. 34/2026.
Abstract: ABSTRACT AUTONOMOUS MULTI-AGENT RETRIEVAL-AUGMENTED GENERATION SYSTEM FOR SCIENTIFIC DOCUMENT UNDERSTANDING The present invention discloses an autonomous multi-agent retrieval-augmented generation system (100) for scientific document understanding. The system (100) comprises a data ingestion and preprocessing module (110), a multi-agent orchestration module (120) with router agent unit (122), planning agent unit (124), and reasoning agent unit (126) for autonomous query decomposition, a retrieval module (130) combining sparse vector retrieval unit (132) and dense vector retrieval unit (134) with reranking unit (136), a quality control module (140) with relevance assessment unit (142), factual verification unit (144), and internal critic unit (146) using multiple LLMs, a response generation module (150) with citation attribution unit (156), and a vector database module (170). The system achieves retrieval F1-score of 0.86 and overall response quality score of 0.92 across healthcare, education, legal, and scientific research domains. [
Disclaimer: Curated by HT Syndication.
