Radius Intelligence (www.radius.com) empowers Data Science to deliver an unique marketing intelligence platform used by hundreds of US companies. At Radius we have moved our entire data processing platform from Hadoop to Spark and this presentation will discuss how data scientists, data engineers and product managers come together to explore data and build new data processing and predictive models on top of our database of tens of millions of US businesses. The presentation will explain how Spark is used to deliver high speed matching across hundreds of millions of records leveraging Scala / Spark for data processing and how MLLib machine learning libraries are used to resolve and impute values for the Index.