This study presents a comprehensive framework for designing data-driven automation architectures that enhance scalability, adaptability, and intelligence in enterprise systems. The research addresses the persistent challenge of integrating automation logic with heterogeneous enterprise environments while maintaining real-time responsiveness and operational transparency. The purpose of this study is to develop a scalable architecture that utilizes structured and unstructured data to optimize automation decisions, resource allocation, and system governance. Employing a mixed-methods approach, the research combines quantitative performance analysis from simulated enterprise workloads with qualitative insights from automation architects and IT process engineers. The proposed architecture leverages a multi-layered orchestration model spanning data ingestion, analytics-driven decision engines, and feedback-based adaptation to demonstrate measurable improvements in process efficiency and governance control. Empirical results show an average 24 percent improvement in automation throughput and a 19 percent reduction in execution latency compared with rule-based frameworks. The study introduces the concept of continuous intelligence, in which automation frameworks evolve through real-time data assimilation and feedback learning. By embedding analytical intelligence within process automation, enterprises can achieve a self-adaptive ecosystem capable of anticipating operational anomalies and aligning automation outcomes with strategic business goals. The findings contribute to both theory and practice by defining a blueprint for next-generation enterprise automation that integrates data-centric design, predictive decision-making, and governance awareness into a unified, scalable framework suitable for digital transformation initiatives.