Agreed, airflow and dbt have literally nothing to do with the size of the data and can be useful, or overkill, at any size. Dbt just templates the query strings we use to query the data and airflow just schedules when we query the data and what we do next. The fact that you can fit the whole dataset in duckdb without issue is kind of separate to these tools, we still need to be organised about how and when we query it.