English
How to develop automated a Power BI and data analytics platform on Azure
AnalyticsCreator automates the development of Azure-based data analytics platforms by generating SQL Server, Azure Data Factory, and Power BI models from a single metadata-driven design. The webinar demonstrates how data engineers can build historized data warehouses, generate ETL pipelines, deploy Power BI datasets, and automate data modeling significantly faster than traditional manual approaches.
Questions
- How does AnalyticsCreator automate Power BI and Azure analytics platform development?
- Can AnalyticsCreator generate Azure Data Factory pipelines automatically?
- How does AnalyticsCreator handle historization and slowly changing dimensions?
- Does AnalyticsCreator support Power BI semantic model generation?
- Can AnalyticsCreator work with Azure SQL Server and SQL Server on-premises?
- How does AnalyticsCreator support near real-time analytics workloads?
Key Takeaways
- AnalyticsCreator is a metadata-driven automation application for building Azure analytics platforms.
- The application generates SQL Server code, ETL processes, Power BI models, and deployment packages automatically.
- Azure Data Factory ARM templates can be generated directly from AnalyticsCreator metadata.
- Historization and SCD processing are automated within the persisted staging layer.
- Power BI semantic models can be generated directly from the data mart layer.
- The repository architecture is open and extensible.
- AnalyticsCreator is a design-time-only application with no runtime dependency.
- Near real-time processing scenarios are supported through Microsoft technologies.
- Data lineage visualization helps developers understand dependencies across large data warehouse projects.
- The generated deployment packages can target Azure SQL Server or on-premises SQL Server environments.
Transcript
With AnalyticsCreator, you design a holistic data model across your entire data landscape. The model covers the flow from source systems through the data warehouse, data marts, tabular or OLAP models, and finally to reporting tools such as Power BI.
Based on this design, AnalyticsCreator automatically generates the source code, SQL code, ETL jobs, and related objects. It is a pure design-time application with no runtime component and no vendor lock-in. Once the generated code has been deployed, AnalyticsCreator is no longer required to run the solution.
Several customers have already achieved significant results with AnalyticsCreator. One customer reported saving 80% of time and costs in a data warehouse project. mymuesli needed only four consulting days to get started and later built its own internal data warehouse capabilities.
Another customer reported being 20 times faster than originally estimated. These examples show how automation can reduce effort, accelerate delivery, and help teams focus on higher-value work.
The automation process starts by connecting your data sources. AnalyticsCreator then reads and manages the metadata layer, which becomes the foundation for the model.
Next, the intelligent wizard creates a draft data warehouse model. After that, you refine the model by adding business logic, transformations, calculations, and other requirements. Finally, AnalyticsCreator generates and deploys the source code to the target environment.
AnalyticsCreator is focused on the Microsoft environment and supports Azure SQL, SQL Server, SSIS, Azure Data Factory, and Power BI.
It can also generate models for other reporting technologies such as Qlik and Tableau. This allows teams to manage the full analytical architecture from a central metadata-driven design.
The build process starts by creating a connector to the source system. After that, the Data Warehouse Wizard is launched and uses the source metadata to generate the required warehouse structure.
AnalyticsCreator automatically creates the layers, including staging, persisted staging, core, and data mart. It also generates dimensions, facts, and historization logic based on the selected configuration.
The persisted staging layer is used to historize data. AnalyticsCreator supports slowly changing dimensions and allows historization behavior to be configured per column.
Transformations can be created using predefined logic, custom SQL, and reusable macros. This makes it easier to standardize recurring transformation patterns while still allowing flexibility for specific business logic.
AnalyticsCreator creates fact transformations based on the selected source tables and relationships. Measures such as quantity, total amount, and discount can be added directly to the model.
Calendar dimensions are generated automatically and can be connected to date fields, allowing facts to be analyzed by time without manual calendar modeling.
For deployment, AnalyticsCreator generates a deployment package. The DACPAC is used to deploy the database structure to Azure SQL.
AnalyticsCreator can also generate Azure Data Factory ARM templates and automatically prepare the Power BI dataset. This allows the modeled warehouse to be deployed into the Azure environment with minimal manual effort.
The generated ARM template creates the required Azure Data Factory pipelines. Workflow packages orchestrate the import, historization, and transformation steps.
Depending on the target environment, AnalyticsCreator can generate either SSIS packages or Azure Data Factory pipelines, giving teams flexibility across on-premises, cloud, or hybrid architectures.
The generated Power BI dataset already contains the required dimensions, measures, and relationships. All metadata is prepared from the warehouse model.
This means users can immediately start building reports and dashboards without manually recreating the semantic model.
Some customers import data every five or ten minutes, depending on their requirements and infrastructure.
Power BI can support thousands of simultaneous users when the model and environment are designed properly. AnalyticsCreator supports this by generating optimized Microsoft code and preparing the required structures for scalable analytical solutions.
