Microsoft Azure is a cloud platform that comprises over 200 products and services enabling the deployment of new solutions. But how can you navigate this ocean without a guide? Let’s look at the necessary services to find the path to your goal.
Azure is Microsoft’s public cloud platform, offering numerous cloud computing services. These include resources for computing, storage, memory, data transmission and network interconnection, analytics, intelligence/machine learning, security and identity management, monitoring and management, and application development.
The services provided by Microsoft Azure can be classified into three areas, depending on the delivery model adopted: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS).
Although you can start for free and have many services without any cost for 12 months, each service involves payment based on consumption, with cost determination methods specific to each service. However, around sixty services are always free.
Understanding which of these services are useful for creating a new architecture is not always simple and intuitive. To make it clear how these services can be beneficial, we will contextualize them in use cases for implementing a new Business Intelligence architecture.
Which Azure Services Are Necessary for Implementing a New BI Architecture?
Azure Data Factory
Azure Data Factory is a managed cloud service for data Extract-Transform-Load (ETL) / Extract-Load-Transform (ELT) operations. This serverless scale-out service offers a code-less visual user interface that is intuitive and easy to learn. It also features a unified console for monitoring and managing connections.
Azure SQL DB
Azure SQL DB is a PaaS relational database service designed for the cloud, always updated to the latest version and fully managed. Both Azure SQL DB Single Instance and Azure SQL DB Managed Instance are available. The Single Instance provides isolated databases with guaranteed dedicated computing resources such as memory/storage. Managed Instance can host a collection of non-isolated databases, enabling concurrent use of information from multiple databases and ensuring data availability through replication to other regions. Both options allow dynamic scaling of resources based on workload needs and offer various configuration/pricing solutions.
Azure Automation
Azure Automation provides a service for automating and managing assets across all areas and subscriptions of a specific tenant.
Azure Analysis Services
Azure Analysis Services is a fully managed PaaS platform that provides the creation of enterprise-level data models in the cloud. Advanced mashup and modeling features allow combining data from multiple sources, defining metrics, and securing data in a single tabular semantic model. This data model enables users to perform ad hoc data analysis more quickly and easily using analysis tools such as Power BI and Excel.
How to Manage Data Sources?
Managing data sources when implementing a new architecture is one of the first challenges to tackle. We will present two possible solutions, but hybrid solutions are also possible. In the following examples, we will exclusively use Azure SQL DB Single Instance, but all logic can be replicated with a Managed Instance.
If we want to create an architecture that keeps an existing on-premises database active, it would be necessary to set up an instance of Azure Data Factory for data loading, using appropriate pipelines, and an Azure SQL DB (less performant) to host the configuration tables.

However, you might decide to migrate data to the cloud and decommission your physical servers; in this case, it will be necessary to create an instance of Azure Data Factory, as before, connected to two Azure SQL DBs: one database will take on the role of the corporate DWH while the other will still serve for the configuration tables and to host all possible loading logs (as defined in the first scenario). It is not necessary to set up two separate Azure SQL DBs for configuration tables and the DWH; a single Azure SQL DB could be used for both purposes.

Setting Up Azure Analysis Services
Another possible scenario to consider is whether to include the Azure Analysis Services in the architecture. The advantage of this service is that it builds semantic models that allow for updates independent of the data sources and the Power BI datasets they feed. Building tabular models aids in self-analysis and optimizes data performance and usability.
If the introduction of Azure Analysis Services is planned, in addition to the data source management flow previously discussed, it will be necessary to set up Azure Automation to automate the execution of APIs towards Azure Analysis Services and/or Power BI. Azure Analysis Services, with the support of development tools such as Visual Studio, allows for the creation of semantic models that make data available for reporting activities on Power BI.

Logic App: A Free Service for a Better Architecture
Among the numerous services offered by Azure, Azure Logic App is a free service (available with a valid subscription) that allows, for example, the creation of an automatic notification system for completed activities. This enriches the architecture by enabling timely management of crisis situations, such as data update failures.
Logic App is a cloud platform that allows the creation and execution of automated workflows without using code (code-less). By using the visual designer and selecting from pre-built operations, it is possible to quickly create a workflow that integrates and manages apps, data, services, and systems.
If notification functionality is to be introduced, it would be necessary to set up a flow from Azure Data Factory, Azure Automation, and the Azure Logic App service. Through a service mail account (e.g., Office 365), it is possible to automate the sending of emails for multiple scenarios, rather than creating small ad hoc flows to enrich reports and/or semantic model tables.

We realize Business Intelligence & Advanced Analytics solutions to transform simple data into information of freat strategic value.