MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202641122894 A) filed by Koneru Lakshmaiah Education Foundation on October 14, 2026, for Framework For Reduced Llm Calls With Localised Modules And Line Fusion In Generative Chatbots.
Inventors include K. Sri Hari Narayanan; and Dr. Nilu Singh.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: A capability-aware framework, Line Fusion (100), enables localised editing of output generated by a large language model (104) without regenerating it. The output is stored as a document model of typed elements. A user selects a portion, such as a sentence, number, table row or code line, and gives an edit instruction; the selection is anchored by element identifiers, offsets and a hash. A request intelligence engine (120) determines the requested operation and the capabilities required, from language transformation, mathematics, code, tabular processing and knowledge retrieval, and builds and schedules tasks. A root decision layer (130) routes each task to locally executing modules (140), or to a knowledge fusion engine (150) that retrieves from a requested or trusted source, generates grounded replacement content and a citation. A fusion integration layer (170) merges the result at the anchor, validates structure and leaves the rest unchanged, reducing model calls, computation and latency.
Disclaimer: Curated by HT Syndication.
