Tools like Spark, Trino, etc can be pointed at parquet/iceberg/etc files in S3, and they'll let you issue SQL queries against the files directly. Means it integrates out of the box with whatever data tooling is being used in your org already.