This session will briefly introduce the concept of re-usable AI building blocks (AI modules) based on Docker containers and protobuf interface specifications. It will also explain the separation of the configuration part and the execution part by demonstrating the public AI playground. Then several concrete examples of AI pipelines and use cases will be presented. All this is based on the Eclipse Graphene project. The session finishes with an outlook to the upcoming features like reproducibility and benchmarking. At the end, attendees will have good understanding how Graphene can help them with thier AI projects.