Design a machine learning solution: Review

The 'Design a model deployment solution' unit is just a series of decision trees in text form. “If you need X, consider Azure Service Y. If you need Z, consider Service A or B.” I clicked through it all in under an hour, hoping for a sandbox or at least a sample repo to clone. Nothing. You don't build anything, you don't configure anything, you just read a list of product names and their intended use cases. It feels like a PowerPoint deck someone converted to a webpage for an exam cram session. The diagrams showing deployment targets are okay, but you could get the same information from a blog post. Don't expect to come out of this knowing how to actually connect Databricks to Azure DevOps. You'll just know that it's a thing you're supposed to be able to do.