Currently some concepts for a Python data pipeline runner
dag_approaches contains a couple of attempts at building a config-driven DAG pipeline runner.
native_python_example contains a working example that relies on Python to procedurally validate object references at runtime, instead of building a DAG ourselves.