Flink), batch, and file-based vendor feeds (SFTP/managed file transfer), landing data from Product Engineering services into the lakeThe lake/lakehouse — Parquet, Trino, table formats, partitioning, and the physical and analytical modelling underneathThe serving/cache tier — powering customer-facing screeners and analytics at production latency … schema evolution, backfillsStrong on schema registries and contracts (Avro/Protobuf, evolution, breaking-change handling across producers and consumers) — not just dimensional modellingFluent in Parquet, Trino, streaming (Kafka/Flink), and lakehouse platforms (Snowflake/Databricks), and the throughput and cost engineering that comes with scaleProduct and governance instincts ...