Transactional data warehouse analytics

A PostgreSQL data warehouse ingests millions of betting and casino transactions. My work here is on the analytics side: learning the warehouse schema and its ETL patterns, then writing the queries that turn raw aggregated transaction tables into answers.

Coming from a MySQL background, this meant becoming properly fluent in PostgreSQL - CTEs as first-class citizens, different indexing behaviour, different temporary-table semantics - and learning to read an unfamiliar schema the way you'd read unfamiliar code: follow the keys, verify everything.

The day-to-day output is analytical SQL against aggregated transaction data: reconciling figures across layers of the warehouse, validating ETL output, and building queries that finance and operations teams can rely on.

• —Reliable analytical queries over a warehouse I didn't design - documented as I went.
• —A working map of the ETL flow, so anomalies in the numbers can be traced to their source.
• —Solid PostgreSQL skills added alongside MySQL.