Sebastian Uhlig

For twenty years I worked as a solution architect on enterprise data platforms, in German and international projects, across finance, healthcare and logistics, on SAP Business Data Cloud (Datasphere), SAP BW/4HANA, SAP Analytics Cloud, Databricks and the hyperscalers.

In almost every one of those projects the same situation came up. The model was built and the reports were live, and then someone from the business pointed out that a mapping was wrong, a hierarchy had changed or a classification was missing. Each project solved that on its own, with a small app or an upload built under deadline, and those solutions rarely survived the project that paid for them.

The pattern repeated often enough to be worth a product. We built NextTables for SAP BW so that business teams maintain their own tables inside the SAP Business Warehouse system, validated at entry, authorized by row, with every change on record.

When enterprise data moved to cloud platforms, the same situation came back in a new architecture, so we rebuilt NextTables from the ground up. It runs on SAP Business Data Cloud and Databricks today and went cloud-ready in 2026. The company started inside NextLytics and is fully independent now.

The conviction behind both products has held for twenty years: a data platform pays off when business teams drive the data as well as consume it. That question is getting sharper as AI arrives. AI systems depend on the mappings, classifications and rules that only business teams can supply, and once agents write to enterprise data, validation, authorization and traceability have to hold for an agent exactly as they hold for a person.

That is what I write about: how enterprise data actually gets maintained, and what changes when AI joins in.

Off the clock I am on a mountain bike or watching American football.