The Analytics Revolution

CHAPTER 5 Creating an Analytics Platform

The analytics landscape has become increasingly complex in recent times. In today's world of operational analytics, it is no longer as simple as picking a database and an analytics tool. There are many new tools and technologies to consider adding into a modern analytics environment. These technologies include nonrelational platforms such as Hadoop, discovery platforms that support both relational and nonrelational data and processing, in-memory analytics, graphic processing unit-based analytics, complex event processing, and embedded analytics libraries. We'll talk about each of these.

Over time, further integration will make analytics environments more and more seamless and simple to use. Today, however, it is necessary to deal with a range of components within an analytics platform. The key is to make sure that the platform as a whole is set up so that it can support all analytics needs. This means not just the needs that exist right now but the needs that are anticipated over the next several years.

For operational analytics to succeed, it is necessary to stitch the components together differently and more completely than in the past to create a single, unified analytics environment that can scale to handle any type and volume of data for any kind of analysis. This may sound like an impossible goal, but the market is evolving rapidly today and it is already possible. In this chapter, we discuss how to make sense of all the options ...