Mitigating Data Leakage Risks: A Comprehensive Approach to Data Governance in Age of AI

Atharva Rajiv Weginwar, Raghunandan Nuggehalli Ramesha, Jayaprakash Ramsaran, Suyash Bhogawar

Abstract: The use of Large Language Models (LLMs) is in- creasing in various industries, leading to growing concerns about data governance. Although LLMs can handle natural language processing and generation exceptionally well, there are important issues to consider regarding data privacy, security, and compliance. This paper examines the hurdles that LLMs present in data governance in different industries and suggests a solution using microservice architecture customized to each department?s require- ments. By incorporating fine-tuned language models specific to each domain within a microservice structure, companies can reduce the chances of data breaches and improve their data governance processes.

Keywords: Large Language Models (LLMs), Data Gover- nance Challenges, Microservice Architecture, Data Privacy, Data Security, Regulatory Compliance, Data Segregation, Scalability