MUMBAI, India, April 17 -- Intellectual Property India has published a patent application (202641043826 A) filed by KKR & KSR Institute of Technology and Sciences, Guntur, Andhra Pradesh, on April 6, for 'privacy-preserving multi-agent financial intelligence platform with explainable advisory mechanism.'
Inventor(s) include Dr. G. Murali; Mr. Tenali Sujith Kumar; Mr. Sk. Faisal Ahmed; Mr. U. Hemanth; and Mr. P. V. S. Subhash.
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: "The proposed invention is a privacy-sensitive multi-agent financial intelligence platform that will provide users with personalised, explainable financial advice without exposing sensitive user information to potential harm. Its fundamental component is a context governance server that implements the self-consent and privacy choices of the user prior to any AI agent accessing financial data - with field-level masking, temporal aggregation and per-request access control to apply in a manner that does not rely on prompt content and thus resistant to indirect access to extraction attempts. An orchestrator organises a group of specialised agents each tasked with a particular part of the financial analysis: knowledge retrieval, financial computation, risk and compliance assessment, user memory and human-readable explanation. The knowledge layer is a combination of financial document semantic search using vectors and graph-based access of typed entity relationships and allows the system to reason about multiple degrees of separation, such as, the acquisition chain of a portfolio holding being indirectly impacted by a regulatory event.All the recommendations generated by the system are based on evidence found, and they are marked with confidence scores and provenance. Concurrently, all policy choices and actions of agents are logged in an unalterable audit store that can be inspected and reviewed by the regulator post hoc. All of these features combined render the system not only viable as a research prototype but also as the design of responsible, production-scale financial AI."
Disclaimer: Curated by HT Syndication.
