How Automated Deployment works within Business Intelligence

Power BI as a starting point

An OTAP street, a separate test and acceptance environment, version management and review. Preferably all automated. A familiar process for the development of business critical applications, but not when it comes to Business Intelligence. Maybe this is because BI is often not seen as software development or because tooling like Power BI does not support it. At TeamValue we want to change this. Power BI is a great tool to create beautiful and clear reports in a fast and accessible way. Especially if you link a standard way of releasing to it. That's why we have developed a process for all our customers using basic functionality in Azure DevOps. By using this solution, we automate the release pipeline of all reports. And that according to our own way-of-working The Golden Path. This way we ensure a flawless release of all reports, and that in the cloud.

Why version control of BI reports is a must

Different versions. Different laptops. We often see that BI reports are not stored in one central location. Implementing a change is then a time-consuming process. That's why we always use version management. This makes it easier to retrieve and use a dashboard or report quickly, should this be necessary.  

Azure DevOps' Git integration makes it possible to use this automated process for BI. This Git integration not only stores all reports in one place, but also gives the ability to review the work of others. This way the adjustments are checked with an extra review by a colleague (four-eyes principle) but we also limit human errors to a minimum. This way of working is one of our principles of The Golden Path.  

Automated Deployment in Practice

For publishing reports we use the Azure DevOps Pipelines functionality. The API that Power BI makes available can fully automatically deploy reports to workspaces and adjust the connection settings. This means you will never have to manually connect to other data sources again once a report goes from acceptance to production.  

Releasing a new report has never been easier and more secure with the approvals included in the pipeline. Where you might have needed three teams for deploying (development, BI, data delivery), it is now possible to do this yourself or directly together with the business. As a developer or BI specialist you will never again need separate rights to a production environment. There are no complex processes anymore and there is less dependence. How is this possible? By using a pipeline with a service account.  

Release Management and One Button Deployment with Azure DevOps

As indicated earlier, Azure DevOps offers, next to full Automated Deployment, also the possibility to set up the process properly by means of releases. Using clear release stages and approvals, the status of a report is completely transparent. With just one push on the button, One Button Deploy, the report can be released. This ultimately ensures a transparent process and makes it possible to give the end user more control over his own reports.  

Unique in the market

With the help of Azure DevOps and Power BI we can now release reports fully automated. This reduces the chance of errors and limits security risks. It also offers transparency over the entire process and gives stakeholders and the business control over the reports. The fact that we combine deployment and management is unique in the market. Whoever follows the Golden Path and thus properly organizes the process from acceptance to production, applies version and release management, is assured of reliable, secure and flexible reporting. Automated Deployment is a solution for both the tech and the process side.  

Need help with the first steps of your reporting? Are you looking for the road that leads you to Automated Deployment within Business Intelligence? Or do you just have a nice technical question? Then get in touch with us.

Also curious about the first step towards a modern data warehouse? Read the blog about Power BI Dataflows here.