Clojure and Large Data: Working the Middle ground between RAM and a Cluster

We all want fast answers, even as our data sets grow in size. Simple techniques stop working when our data no longer fits in a single process. In this talk I’d like to share a few of the techniques I’ve learned that stretch what you can do on a single machine. I’ll cover taking random samples from large data sets with stream sampling and identifying (probable) duplicates when the set is too large to fit in memory.