MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202621017403 A) filed by Dr. Nilesh Sable; Mrs. Archana Burujwale; Tejas Terdale; Samruddhi Shinde; Vaidehi Sarda; and Avni Selote, Pune, Maharashtra, on Feb. 17, for 'verification-driven dual-stream retrieval system for statutory legal auditing.'
Inventor(s) include Dr. Nilesh Sable; Mrs. Archana Burujwale; Tejas Terdale; Samruddhi Shinde; Vaidehi Sarda; and Avni Selote.
The application for the patent was published on April 17, under issue no. 16/2026.
According to the abstract released by the Intellectual Property India: "Statutory auditing requires a level of precision that conventional semantic search-based legal systems often fail to deliver. Many modern legal retrieval frameworks assume that neural embeddings or hybrid retrieval methods are universally effective across domains, including statutory law. The present invention introduces a verification-oriented statutory auditing system designed to evaluate retrieval accuracy at the level of exact legal sections and clauses. The system employs parallel lexical and semantic retrieval pipelines to independently process statutory queries. A transparent rank-fusion mechanism is used to reconcile retrieval results without presuming semantic dominance. Statutory documents are grounded through structured chunking and explicit metadata annotation to ensure traceable legal citations. A controlled synthetic benchmarking process generates unambiguous queries with clear ground-truth references. The invention evaluates retrieval performance using established information-retrieval metrics relevant to legal decision-making. All system components operate entirely on local hardware, ensuring privacy and regulatory compliance. The invention demonstrates that lexical retrieval remains critical for high-precision statutory auditing while challenging prevailing assumptions in legal AI system design."
Disclaimer: Curated by HT Syndication.
