Track: Data Management
- Mittwoch
25.06.
How successful is our content on traditional television compared to our streaming platform, Joyn? And how do we ensure that as many viewers as possible tune in? These are just a few examples of key business questions that we answer using our Data Warehouse.
In this presentation, we will take a closer look at our BI transformation and demonstrate how we have successfully implemented "agility" in the context of data warehousing within a multi-team setup. In addition to innovative approaches to…
The session will discuss and present how ML Models can be integrated with MLFlow for Tracking, Deployment, Versioning, etc. It will cover in detail how ML Models and its artifacts are logged to SQL Server and MLFlow for deploying efficiently. We will also see how a model can be dockerized successfully and pushed to DockerHub so that end users can use it as a Docker Container.
All this will be done through a live demo of a project of which all the artifacts and final model will be logged to the…
To implement the Once-Only principle, we are required to introduce a strictly register-centered data management for the entire administration. The aim is to achieve broad data interoperability between registers and information systems. This talk will illustrate our ongoing, comprehensive change process from an organizationally grown data management to a modern, register-centered, standardized data management based on harmonization. We highlight the necessary building blocks that facilitate such…