Lightweight, open-source Large Language Models (LLMs) are rapidly transforming the way businesses adopt artificial intelligence. Some of these models can be run offline directly on laptops and personal computers, offering cost-effective, free, and privacy-focused alternatives to cloud-dependent AI solutions, thereby making advanced AI capabilities accessible even to small and mid-sized organizations, said CA Dhananjay Gokhale, Business Intelligence professional at Perficient Inc.
He was addressing participants during the AI Marathon Series organized by the Vidarbha Management Association (VMA) at Chitnavis Center, Civil Lines, Nagpur. The session highlighted the rapidly growing role of lightweight, open-source LLMs in modern business environments. The Project Director for the session was Mahendra Girdhar.
CA Gokhale further elaborated on the introduction of an open-source platform that enables users to run LLMs locally while ensuring complete data security and privacy. Participants actively engaged in hands-on exercises, including generating letters and documents, summarizing content, and converting scanned images into editable text using lightweight LLMs.
A key highlight of the session was the live demonstration of integrating Tally Prime with Claude using Model Context Protocol (MCP) via an open-source connector. Attendees observed how simple natural-language prompts could fetch ERP data, generate dashboards, and derive deep financial and business insights—tasks that are otherwise time-consuming when done manually. These insights were seamlessly exported into Excel and Word formats.
👉 Click here to read the latest Gujarat news on TheLiveAhmedabad.com