The biggest challenge for the big data industry is to bridge the worlds of statisticians and cloud based big data platforms. The missing layer in the big data stack is an interactive software platform that allows large groups of data analysts to build, validate and deploy complex models. It goes without saying that the building blocks of this stack are Amazon Web Services, Hadoop and R.

Suppose we gave the modelers direct access to Hadoop and Amazon Web Services? The data analysts would have to learn distributed programming frameworks like MapReduce and R. Is there an alternative to giving each statistician a crash course in MapReduce programming?

The MarketShare engineering team has built a platform which enables data analysts to build comprehensive and sophisticated models and allows them to orchestrate a cloud based analytic workflow. The platform helps the analysts build, validate and deploy models in an automated yet customizable manner. The end user is a marketer who accesses these models through intuitive cloud based applications. The marketer uses the applications to answer critical questions regarding marketing resource mix, attribution and dynamic pricing. Through this, MarketShare provides its customers a truer view of marketing's impact on the business, and better guidance on how to act moving forward.

In the next few posts, we will walk our readers through the various components of this platform.