I would strongly recommend reconsidering the suggestion that SQL serve as a data engineering pipeline default. We use SQL heavily at our organization, and use

I would strongly recommend reconsidering the suggestion that SQL serve as a data engineering pipeline default. We use SQL heavily at our organization, and use it for broad, significant ELT workloads. Its use does help to interface across multiple departments to provide business intelligence across our organization. It does serve as a sort of data lingua franca and touchstone for analyst and engineer interactions. In essence, we use SQL, largely via DBT, where appropriate. However, everywhere else, which comprises a significant collection of data pipeline services, uniform where possible but definitely heterogenous, we default to Clojure; we also utilize Python where more appropriate. Of these 3 languages SQL clearly has the least generality, least testability, least API integration capability, and definitely the most awkward data processing capability. I feel it is naive to suggest it could serve as a default language for data engineering pipelines, particularly given the need to interact with cloud services and third parties.