Services like Google’s Big Query, AWS’s IoT Gateway, and Azure’s Analysis Ser-
vices are tremendous accelerators for developers. All of the complex IT plumb-
ing is just there running and scaling on demand. These services are also fully
integrated with the platform, thus reducing ugly integration nightmares such
as trying to cobble together and manage a dozen different point solutions to do
the same job on-prem.
Three Popular Buzzwords: Big Data, Machine Learning &
Artificial Intelligence
All of the big three public cloud providers (Amazon, Google, Microsoft) have done
an excellent job of providing developers with simple APIs that abstract the com-
plexity involved in installing and integrating all of the underlying technology
stacks. Take a look at the complexities in implementing a big data solution.
Figure 1: Sample Big Data Architecture on AWS
This is a very complex architecture that is required to meet all of the business,
security and regulatory compliance requirements of a typical large enterprise.
But what you may not realize is that each icon on this diagram is a managed
service. Each managed service is an abstraction of many underlying technolo-
gies. A company trying to build this in the DIY model would spend weeks or
even months to implement this solution, not to mention a small fortune. Add up
all the icons and you quickly see that it would be easy for a company to “go
dark” on the business for 12-18 months just getting the infrastructure vendor
products ready to write the first line of application code.
That is before we even get to machine learning and artificial intelligence. Pre-
viously, a company would need to hire an army of PhDs to implement the
underlying technologies and train the models required to teach the systems to
learn patterns and discover insights. Once again, the cloud providers are
abstracting away all of these complexities and allowing mere mortals to use
machine learning and artificial intelligence APIs to create business value. I was
52 | THE DOPPLER | SUMMER 2017