A Cloud-Native Modernization Framework for Enterprise Analytics and AI Platforms on Microsoft Azure

Enterprise analytics and AI platforms often begin on-premises, where teams retain control over hardware but face growing constraints in scalability, environment consistency, and deployment automation. This paper presents a six-phase cloud-native modernization framework for migrating and re-architecting such platforms on Microsoft Azure. The framework spans legacy assessment, workload classification and managed-service mapping, target architecture and security design, Infrastructure as Code (IaC) and landing zones, CI/CD and operational automation, and validation with pattern extraction. It is instantiated through an industrial case study in which a platform running on five physical servers — comprising SQL databases, ETL pipelines, machine learning workloads, a web application, notification workflows, and a shared network file path — is transformed into a managed, automated Azure architecture built on Azure SQL Managed Instance, Azure Data Factory, Azure Machine Learning, Azure Functions, Azure App Service, Azure Logic Apps, and Azure Blob Storage, provisioned with Terraform and deployed through GitHub Actions CI/CD. The evaluation reports quantitative outcomes — full decommissioning of on-premises servers, migration of four major workload categories to managed services, standardization of three environments, and implementation of multiple CI/CD pipelines — together with qualitative gains in scalability, deployment confidence, and reduced configuration drift. The paper contributes a reusable modernization framework, a detailed implementation case study, and practical patterns that bridge cloud migration, IaC/DevOps, MLOps, and serverless inference literature for practitioners undertaking similar transformations.