What pg_lake, pg_duckdb, and pg_mooncake do at the Iceberg level, and honest thresholds for when Postgres is enough.
How Avro, Protobuf, and JSON Schema evolve through a registry, and how that maps to the schema evolution rules of Iceberg.
How the spec-defined bucket transform lets engines skip the shuffle in joins, and how to set it up and keep it engaged in Spark.
What synthetic data in a lakehouse is for, the generation methods, how to preserve fidelity, and where synthetic tables belong.
