To do this, I think I need to know:

• Can you trigger a pre-existing Pipeline via the ClearML REST API? I'd want to have a Lambda function trigger the Pipeline for a batch without needing to have all the Pipeline code in the lambda function. Something like curl -u '<clearml credetials>' None ,...
• [probably a big ask] If the pipeline succeeds/fails, can ClearML emit an event that I can react to? Like maybe with a webhook or something? That way an event-driven system could place items to inference on a "retry queue" or a "success queue" and notify accordingly.
• If (2) is not possible: is there a way to add an "try/except" component to a Pipeline? (Something that gets run if any of the previous steps fail). That would be less ideal, because our pipelines would have to be "aware" of our AWS-based queue-ing system, but at least we could react to failing steps and use Python code in the ClearML pipeline to publish the inputs of the failed pipeline to a retry queue. I worry this method would be more flaky.