Most data engineering teams still work in a translation loop. A business team asks for a churn model, a risk view or a ...