Articles

Data and artificial intelligence (AI) models lack the governance and protective mechanisms that traditional applications enjoy through software engineering (DevOps) practices in industry. The complexities of the enterprise CI/CD pipelines, which span across source control, continuous integration, continuous delivery deployments, testing, monitoring for observability, blocking and telemetry for production, enterprise dashboards with pipelines and project health, for AI models and Data in Enterprise environments is presented. Data at Rest, Metadata and Smart data through automation that provide lineage and tracking automation for ML models are discussed. Cloud DevOps ecosystem built with Azure DevOps repository and Pipelines with extensions, which enables orchestration creation and data pipeline YAML templates that embrace best CI/CD practices guide is discussed with detailed indices for enterprise reference. The enterprise ecosystem provides seamless package management integrated with other Azure DevOps extensions and third-party services availability. The pipelines provide the ability to deploy to any environments read from the pipeline YAML config. The development of pipeline YAML is driven as per the extensions installed and project structure for easy consumption by developers with integrated documentation and additional metadata to help in easy maintainability. These pipelines templates can be used to build, test, and publish data artifacts in enterprise data pipeline, enterprise database automation.