Page 177 - Building Big Data Applications

176 Building Big Data Applications

Storyboard additions
Storyboard modifications
In the world driven by big data and internet of things, we need this approach to
provide a focus, a starting point and ensure a successful completion of the process. In
the earlier chapters where we discussed several examples, the success came as the
process optimized, but the early on approach will make a big difference and avoid the
classic programmer burnout and failure, which we have seen in the industry many times.
Use cases are the most important piece of the storyboard and the entire application is
delivered based on acceptance by the end users. Think of CERN and the experiments to
find the Higgs boson particle, if the use case was not specific to what was being
experimented, the teams over the years would have lost sight of the project and the
applications developed, maintained, and retired through the lifecycle of the experiments
would have been failures. The use case needs to be developed with senior leaders who
can provide the best in class requirements and also align the outcomes expected from
the use case. This is the key to getting the best possible budgets as the use case will be
excellent to the enterprise and will be adopted for multiple research projects. An
example here is the use case for home stay and study with customers by Procter and
Gamble, which resulted in phenomenal success for the research inputs to the “Tide”
washing detergent project. If you read “The Game Changer” by A.G.Lafley, he writes in
detail the involvement of the executives and the research teams to bring success to the
customer, which results in profit for the enterprise. Another research project that can be
looked at is the “largest form of cancer and its treatment”, yes we are talking about breast
cancer, while there are 1462 drugs in the market for treating different issues from the
cancer, there are two pharmaceuticals who are on the brink of a good breakthrough,
Pfizer and Roche. These giants have a use case for the research and have used the
platforms for the data very well to keep focused on outcomes, which have proven to be
the tipping point. These use cases will be presented for approvals of the drugs and they
are the same as written by the enterprise when they started the process.
Research projects are the backbone for building applications in any enterprise. The
old method of creating a roadmap and then delivering business intelligence and ana-
lytics are not in vogue. We need to start with defining the outcome needed, for example if
the outcome is to read customer sentiment or customer click behavior, we do not need a
roadmap but a research project to have formulas to classify the customer based on the
outcome and run instant study of the customers as they browse the store on the internet.
This research outcome will lead us to a use case, and that will work its way to a story-
board which then leads us to delivering the application. We need to be different in the
innovation processing the new world, where we need to experiment and validate, this is
why we recommend that all research teams whether in enterprise or university use the
data swamp, a raw data layer with dirty data, which is easy to work with and will not
interrupt the run of the business process. In drug research or other research programs,
this layer is needed to collect the intermediate results for each execution and rerun the