Dagster
Modern data orchestrator where "assets" (datasets, dashboards, ML models) are first-class objects — not just tasks as in Airflow.
Elementl (2018) by Nick Schrock (ex-Facebook GraphQL). "Software-defined assets" declare which data a pipeline should produce; Dagster figures out the dependencies. Built in: type checks, asset lineage, partitions, observability, local dev environment.
Clearly better DX than Airflow for new projects. Sweet spot: teams that take data engineering seriously and want tests + types. Competitors: Airflow (mainstream), Prefect (lighter), Mage.