Let’s imagine I’m building a pipeline with five consecutive steps, where some of the steps are non ML/DL based. Using ClearML I run a lot of experiments to find the right pipeline configuration. After I found the right algorithms and parameters for my pipeline steps, I want to use the pipeline in production. Since I use multiple steps and have non ML/DL steps, I can’t use clearml-serve or any other model deployment. Is there a way to run all pipeline steps, not in isolation but consecutive in the same environment? For a production environment, it wouldn’t work to spin up a ClearML Agent per step and wait for pulling code and uploading artifacts. Just too slow! Basically, I’m looking for something that merges multiple Tasks together and allows local artifact passing without the ClearML Server in-between.