Go internals: memory/heap management and the garbage collector, synchronization primitives (Mutex/RWMutex, atomics, compare-and-swap), slices vs maps and their complexity, channels and worker pools, non-blocking reads with select, data races. Databases: transactions, isolation levels and MVCC in PostgreSQL, indexes and partitioning, Redis, Kafka (partitions and consumer groups). Domain exercises: data structures to process a quotes feed and trigger stop-loss/take-profit orders, and enforcing a "max active trades per user" limit (optimistic vs pessimistic locking). Final round: how to test hypotheses/MVPs without breaking core functionality, and experience using AI tools in the daily workflow.