Abstract
Large enterprises generate vast volumes of heterogeneous data, making timely and accurate decision-making increasingly complex using traditional analytics and reporting systems. This study proposes an AI agent-enabled predictive analytics framework integrated with intelligent dashboard automation to support real-time decision intelligence. The methodology combines machine learning-based predictive models, autonomous AI agents, and interactive dashboards within a unified architecture that enables continuous data ingestion, real-time model adaptation, and automated operational actions. The framework is evaluated across multiple enterprise use cases, including sales forecasting, demand prediction, customer segmentation, and inventory automation. Experimental results show strong predictive performance, with neural network models achieving up to 91 % accuracy, while AI agents enable rapid anomaly detection and automated responses with success rates exceeding 90 %. User interaction analysis further indicates high dashboard engagement and improved decision efficiency. The findings demonstrate that integrating AI agents with predictive analytics and intelligent dashboards significantly enhances operational agility, decision accuracy, and execution speed, offering a scalable and practical solution for real-time decision intelligence in large enterprises.