Ontology Driven Data Quality Management for Agentic Ai Systems: A Framework for Intelligent Decision-Making in Enterprise Environments

Abstract

The rapid adoption of agentic artificial intelligence systems in enterprise environments has exposed a critical vulnerability: poor data quality undermines AI decision-making at scale. As organizations transition from experimental AI deployments to production-grade agentic systems, the "garbage in, disaster out" phenomenon has emerged as a primary barrier to value realization. This paper investigates the convergence of ontology-based knowledge representation, data quality management, and agentic AI architectures to propose a novel framework for intelligent decision-making in enterprise contexts. Through a systematic review of contemporary literature and industry developments from 2024 to 2026, this research identifies that traditional data quality approaches, designed for business intelligence workflows, prove inadequate for autonomous AI agents operating across distributed data ecosystems. The proposed framework integrates semantic ontology layers with automated data quality monitoring, enabling agents to interpret context, resolve ambiguities, and maintain decision integrity. Findings indicate that organizations implementing ontology-driven data quality management achieve measurable improvements in AI decision accuracy while reducing governance overhead. This paper contributes a conceptual architecture and implementation guidelines relevant to both academic researchers and industry practitioners navigating the agentic AI transition.