I expected this to be about using Athena through the AWS SDK to actually query CSVs on S3 using SQL

https://docs.aws.amazon.com/athena/latest/ug/csv-serde.html

But your point stands, sometimes a simple file is a better choice than infrastructure.