MUMBAI, India, Jan. 2 -- Intellectual Property India has published a patent application (202541122168 A) filed by Vellore Institute Of Technology, Vellore, Tamil Nadu, on Dec. 4, 2025, for 'graph-based retrieval-augmented generation system for financial document analysis.'
Inventor(s) include Dr. J Vijayashree; Akshith Thotanthilla; Pratyush Tyagi; and Devayan Das.
The application for the patent was published on Jan. 2, under issue no. 01/2026.
According to the abstract released by the Intellectual Property India: "The present disclosure provides a graph-based retrieval-augmented generation system for financial document analysis. The system includes a document acquisition module that extracts financial documents from SEC EDGAR database, including 10-K annual reports, 10-Q quarterly reports, and 13F institutional investment filings. A text processing module segments documents into text chunks and generates vector embeddings using a pre-trained language model. A graph database module uses Neo4j to store text chunks as nodes in a graph structure, with sequential linked-list arrangement preserving document flow and company nodes connected through hierarchical relationships. A similarity retrieval module receives user queries, generates query embeddings, calculates cosine similarity scores, and retrieves relevant text chunks based on similarity scores and graph relationships. A response generation module integrates retrieved text chunks with user queries and generates natural language responses using a large language model API."
Disclaimer: Curated by HT Syndication.
