Since 2015 once every year Europe’s top data scientists and data experts from industry make its pilgrimage to its most important gathering, the Data Native Conference in Berlin. This place is special: It’s not an ivory tower event, limited to academics; it’s an event where both scientists and industry experts mingle, brainstorm, code … creating the next big trend. It is also hosted in a special location: The Berlin Kühlhaus (link), build in 1901 as Europe’s largest “fridge” and located near the old Anhalter Bahnhof at Gleisdreieck (link) just south of the famous Potsdam Plaza.
Standing right in the middle of Kühlhaus, on the center stage of Data Natives, Prof. Dr. Christoph Schlueter Langdon of the Telekom Data Intelligence Hub together with Nadine Brehm, Sales Leader at IBM Cloud PAK for Data & Big Data. The Data Intelligence Hub is a platform-as-service solution composed of a Workspace, Marketplace and IDS Connector, offering a secure and efficient way to exchange, process and analyze data. In its basic form, it is using Open Source technology. For example, its Workspace provides a Jupyter Notebook based analytics development environment with a selection of curated Python libraries. The Hub is also providing premium tools. One key example are products from IBM – as indicated on the display on stage.
Figure: IBM and Telekom Data Intelligence presenting the “data pipeline”
Technology innovators like Deutsche Telekom and IBM are expanding data analytics offerings toward “Data factory” capabilities. A data factory economizes on the refinement of raw data into data that is ready for artificial intelligence applications, so-called AI-ready data (see “Data factories for data products,” link). At IBM this orchestration of data capabilities is consolidated as IBM DataOps (link). The Telekom Data Intelligence Hub is expanding its data pipeline with partners (see figure): Examples include analytics tools from IBM and data governance components based on IDS standards.
Data governance is crucial for the success with data: On one hand, some popular AI methods, such as neural networks, require large data pools (see vertical, horizontal data pools, Data Sandwiches in “Creating data pools,” link), on the other hand, data owners are reluctant to share data into pools because of the fear to lose data rights or data sovereignty. How to solve this dilemma of protecting data sovereignty while enabling data sharing? The International Data Space Association (IDSA) provides the reference architecture model (RAM, link) to enable data sovereignty. And the Telekom Data Intelligence Hub is the first to offer an implementation with the IDS_ready Connector (see “Implementing IDSA,” link).