In this talk, I will present techniques for efficient incremental processing of complex analytical queries, ranging from classical SQL queries to linear algebra programs. Our system, called DBToaster, compiles declarative database queries into high-performance stream processing engines that keep query results (views) fresh at very high update rates. DBToaster uses a recursive query compilation algorithm that materializes a supporting set of higher-order delta views to achieve a substantially lower view maintenance cost. Our implementation supports batched processing in local and distributed environments and can deliver up to 5 orders of magnitude better performance than existing DBMS and stream engines.