Where DevOps tries to improve the collaboration between developers and IT pros, the Data Scientists now also have to find their way with the developers and IT pros. In doing so, the concept of DevOps can also be extended very well to the topic of AI - MLOps (DataScience Ops).
This talk will look at how Data Scientists work and how they can collaborate with Developers and IT Pros. Questions answered include:
- What flow is mapped with MLOps?
- Where do I store training code?
- How can I manage my models (versioning/history)?
- How does monitoring work?
- Can I also automate - how do I get that into production?
Through an exemplary flow, I'll address the answers and demonstrate which tools, such as Git, Azure ML Service and GitHub enable implementation.
By the end of the session, attendees will understand why MLOps is important and how to use the tools properly to support a smooth flow.

I am also frequently on the road with this topic in my job and see a constant need to talk about this topic. In companies, I often see a high level of interest in introducing MLOps and getting to know the corresponding tools.
The experiences from the DevOps Journey and the possibilities in the field of AI should be used by the DataScientists but also by the software developers in the AI environment. I believe I can offer something for everyone here.