English
Build an Azure DWH from a SAP Source with PBI on top in just 60 Minutes
Questions
- How can AnalyticsCreator automate SAP data warehousing?
- Can AnalyticsCreator build a data warehouse in 60 minutes?
- What modeling approaches does AnalyticsCreator support?
- Can AnalyticsCreator export to Snowflake or SAP HANA?
- How are transformations handled in AnalyticsCreator?
- Does AnalyticsCreator support metadata-driven development?
Key Takeaways
- AnalyticsCreator is a pure design-time application
- Generates full source code with no runtime dependency
- Supports SAP, SQL Server, Azure, Power BI, and SSIS
- Data warehouse structures are generated automatically
- Historization is fully automated
- Metadata-driven modeling accelerates development
- Supports Data Vault, Kimball, and hybrid approaches
- Supports Data Vault, Kimball, and hybrid approaches
- Power BI models can be generated automatically
- Large SAP projects can be completed significantly faster
- Open repository allows custom extensions and add-ons
Transcript
Hello everybody, and welcome to this event. Today, we will show you how AnalyticsCreator can create a data warehouse from an SAP source in 60 minutes, and how the data can then be accessed and analyzed with Power BI.
My name is Peter Smoley. I am the CEO of AnalyticsCreator and one of the founders. AnalyticsCreator is a Munich-based company, with all development based in Bavaria. The product has now been used at customer sites for more than 15 years.
In a nutshell, AnalyticsCreator is a pure design-time tool. This means the outcome is pure source code that can be used directly in your own environment.
The key component of AnalyticsCreator is its comprehensive data modeling toolkit. It includes best practices and suggestions that help you design your model more efficiently. You can choose Data Vault modeling, Kimball modeling, or our special mixed approach. You are also completely free to develop your own approach.
AnalyticsCreator creates the full data warehouse, including the processes needed to load data from the source into the data warehouse, the cloud, and analytical models.
AnalyticsCreator supports deployments to Azure environments and local SQL Server environments. It can create SSIS packages, Analysis Services packages, tabular models, Azure Data Factory pipelines, and Synapse environments. It can also deploy Power BI models, Tableau models, and Qlik models.
It is also possible to export a data warehouse to target databases such as Snowflake, Amazon, and others.
One of our customers reported saving 80% of costs when creating a data warehouse for a specific IT controlling project using AnalyticsCreator.
Another customer, mymuesli, built their complete data warehouse environment themselves after onboarding with AnalyticsCreator.
Another SAP customer reported being 20 times faster than originally estimated when creating their integration layer.
AnalyticsCreator is a pure design-time tool, meaning there is no runtime component in your environment. The generated code is independent and can be used freely.
The model is stored in an open repository, which customers can extend with their own add-ons.
We are also working on SAP HANA support, allowing customers to generate either Microsoft-stack or SAP HANA-based warehouses from the same model.
Let’s create a new data warehouse project.
First, we add a connector to our SAP source system. AnalyticsCreator supports SQL Server, Oracle, ODBC, text files, Azure Blob Storage, SAP, and more.
In this case, we use a metadata connector because we do not have direct access to the SAP system.
The metadata connector contains SAP FI and Controlling tables, such as booking headers, booking positions, accounts, customers, vendors, and more.
Now we start the Data Warehouse Wizard. The wizard automatically creates a draft version of the data warehouse.
We select booking headers, booking positions, customers, vendors, accounts, company codes, document types, and text tables.
We import them, historize them, create dimensions, and create a fact transformation based on the position table.
The wizard generates the model automatically.
We now see the source layer, staging layer, persisted staging layer, core layer, and data mart layer.
The persisted staging layer stores historized data. Additional valid-from and valid-to columns are created automatically.
Historization packages are also generated automatically. We can configure SCD Type 1 and SCD Type 2 historization, and define filters, variables, and calculated columns.
Transformations in the core layer are usually SQL views.
AnalyticsCreator supports predefined transformations such as trimming, type conversion, and null handling.
Macros allow reusable SQL snippets.
Fact transformations combine data from multiple tables, and transformations can be materialized into persisted tables.
The data mart layer exposes dimensions and facts for Power BI and other reporting tools.
We can also define measures directly in AnalyticsCreator.
Now we deploy the model.
AnalyticsCreator generates a Visual Studio deployment package.
We can deploy Azure SQL databases, SSIS packages, Azure Data Factory pipelines, and Power BI tabular models.
The workflow package automatically orchestrates all ETL execution steps.
The generated Power BI model contains facts, dimensions, and measures, so we can immediately start building reports.
The questions covered topics such as DAX queries, compression, partitioning, metadata annotations, Azure deployment, and Power BI integration.
AnalyticsCreator also supports lineage analysis and filtering for very large data warehouse projects.
The session concludes with information about the BARC Data Management Survey, architecture audits, and the AnalyticsCreator free trial.
Thank you for joining us.
