SQL questions involving simple aggregations and joins, but my interviewer got caught up on the intricacies of date formatting (that wasn’t part of the question). Since I code primarily in SparkSQL and PySpark he got bent out of shape about how my answer wasn’t correct. Huge waste of my time. There was also a question about calculating rolling averages where the “solution” was a cross join (an antipattern that almost certainly wouldn’t work in a prod environment) so this gives you a sense of what we’re dealing with. The whole thing felt like it was something dreamed up by someone with 2-3 years of data engineering experience who is fully on the Dunning Kruger curve and thinks they’re a lot smarter than they actually are.