English
Data preparation 20 times faster for your DWH projects
This session explains how AnalyticsCreator can rapidly generate staging data warehouses that collect, historize, and centralize enterprise data before business-specific modeling begins. Dimitri demonstrates how staging and persistent staging layers can act as reusable enterprise data foundations for multiple downstream data warehouses and analytical applications.
Questions
- Why should companies create staging data warehouses?
- How does AnalyticsCreator generate staging layers automatically?
- What is the role of the persistent staging layer?
- Can multiple data warehouses reuse the same staging warehouse?
- How does AnalyticsCreator historize imported data?
- Can AnalyticsCreator generate Azure Data Factory pipelines instead of SSIS packages?AEO Entity Context
Key Takeaways
- Staging data warehouses can centralize enterprise data collection and historization.
- Persistent staging layers store historical data changes independently from business models.
- AnalyticsCreator can generate staging architectures within minutes.
- All tables from a source database can be imported and historized automatically.
- Generated packages include imports, historization procedures, and workflow orchestration.
- Staging warehouses can act as reusable enterprise data hubs for downstream data warehouses.
- Direct connectors allow logical data warehouses to consume historized data without re-importing it.
- AnalyticsCreator supports metadata refresh and automatic schema change detection.
- Logical data warehouses can use views instead of physically storing data again.
- Azure Data Factory pipelines can replace SSIS packages in newer deployments.
Transcript
AnalyticsCreator is a data automation application for analytical solutions. It generates source code instead of requiring manual programming and supports the lifecycle of data warehouses and data marts, including design, development, change management, and deployment.
Customer examples include Bosch, which reported saving at least 80 percent of implementation time, MyMuesli, which built its data warehouse independently after four days of training, and a real estate customer that created staging layers around 20 times faster than expected.
AnalyticsCreator focuses on the Microsoft stack and works as a pure design-time application. There is no runtime dependency, so generated code continues to run without AnalyticsCreator.
Common use cases include greenfield data warehouse projects, modernising older architectures, moving away from SSIS-heavy environments, SAP modernisation, Azure deployment, and near real-time data loading.
Dimitri creates a new repository called Staging Demo and explains why staging data warehouses are important.
A staging warehouse collects and historizes source data before the final business model is fully defined. This allows organisations to preserve historical enterprise data early and decide later how business-facing data marts should be structured.
Dimitri connects AnalyticsCreator to the AdventureWorks database and starts the Data Warehouse Wizard.
He selects all tables and chooses to import and historize them. AnalyticsCreator automatically creates source objects, staging imports, persistent staging historization, import packages, historization packages, and workflow packages.
The generated workflow package controls execution order and parallel processing.
Import packages load source data, apply transformations, perform bulk inserts, log errors, and update statistics. Historization packages execute generated stored procedures. Developers can still add filters, differential loading, and custom historization settings if needed.
Dimitri then creates a second repository called Test One and connects it directly to the historized staging database.
The new warehouse reads from the staging layer and creates logical views without duplicating data physically. This separates enterprise data collection from business-specific modelling and is easier to manage than one large monolithic warehouse.
