Laughable SQL round administered by a “senior” data engineer with no hands-on experience with Spark or open table formats. Nitpicky questions during a “pair programming” exercise where the assumption is that you know nothing (I’m a long-tenured FAANG employee). Read on if you’re curious, but generally, just stay away.

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. It was a 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.