Architectures for analytical data platforms

In the post-DWH era, a discussion of existing approaches to data architectures is not only essential from an academic perspective, but especially if a specific goal is to be achieved. Not all architectures are useful for all objectives. The common data architectures need to be discussed and categorized: from the data warehouse to the data lakehouse.

On the data side, this means taking a special look at Data storage, data processing, data utilization, data management in comparison.

Technological basics that need to be verified: Separation of compute and storage, schema-on-read vs. schema-on-write and the path from CSV to delta file.

Currently exciting: Technological alternatives: Synapse vs. Microsoft Fabric
What should be done, for example, if legal requirements do not (yet) permit cloud use?

Data Lakehouse onprem? These are important practical topics that we will be discussing here in the near future. Please also take a look at our corresponding and well-attended event on this topic LinkedIn Data Architectures.

The questions and suggestions have shown that the combination of Synapse Analytics, Python and SQL is a future-proof solution for the challenges of modern data architectures.