Enterprise decision-support and business intelligence (BI) systems increasingly demand real-time analytics, multi-tenant scalability, and continuous integration of artificial intelligence models to maintain competitive advantage. Traditional virtualized infrastructure struggles to meet the latency, throughput, and cost-efficiency requirements of modern AI-driven analytics workloads. This paper presents a comprehensive architecture for scalable enterprise BI platforms built on containerized AI workflows orchestrated through Kubernetes on OpenStack infrastructure. Building upon validated container-based frameworks that demonstrate superior performance over virtual machines for multi-tenant AI workloads, this research operationalizes infrastructure-level optimizations into business-layer applications. The proposed architecture integrates continuous integration/continuous deployment (CI/CD) pipelines, GPU-enabled autoscaling, and secure multi-tenant isolation to deliver measurable improvements in decision cycle speed, operational efficiency, and resource utilization. Through analysis of contemporary implementations and empirical evidence from production deployments, this study demonstrates how containerized orchestration translates infrastructure gains, including latency reductions up to 80×, throughput improvements of 2.4×, and cost savings exceeding 40%, into tangible business outcomes across finance, telecommunications, e-commerce, and operations domains. The findings establish a blueprint for enterprise architects seeking to modernize BI infrastructure while maintaining security, compliance, and scalability for heterogeneous analytical workloads.