The ability to manipulate and query very large datasets in realtime is a pressing need for most large data enterprises. Recently, we've seen an explosion of tools such as Impala or Druid, but all of these tools suffer from single points of failure or can't deliver the sub 1 second query times necessary for realtime results. Together we'll explore how to break down these seemingly intractable problems. We'll learn how to build horizontally scalable query engines with Cassandra, capable of sub-second query times across multi-billion row datasets.