The key question: what to do with Big Data, how to deal with it?
Compare it with eating an entire elephant: you will eat it in bits. Every day a tiny bite. Not too large, otherwise you will choke, and never to skip a day – if you want to eat an entire elephant you cannot afford to miss a day. Big Data is basically the same: make it small and tangible, start with a tiny bite and continue evenly.
Don’t get carried away in defining Big Data processes. Don’t be seduced into designing a firm, future-proof, scalable, and flexible target architecture – before you even get started. This would be like “solving world hunger”; too complex to easily solve it, too broad to make it tangible.
Please note: a good design is absolutely necessary. However: don’t let it be your first bite.
The same goes for data-ownership issues or data-governance arguments. Those are topics to discuss, in order to be able to provide answers to the many questions raised. But again: don’t let it be your first bite.
Where to begin?
Let the data speak foritself right as from the start. So: immediately upload that set of data in “any tool”. Goal here is to directly explore these large amounts of data – in order to gain new insights through statistic methods, correlations, visualization and machine learning.
Explore data step-by-step: compose a small team, one that at least consists of a data scientist and a domain expert.
The data scientist is necessary to do his magic with algorithms – to find a correlation, a trend line or an anomaly within the data. The domain expert is necessary to add business-process knowledge that will bring the value of data to life.
And the bites? Research one hypothesis every day, reveal a tiny bit of added-value every day.
And get added value from the business process.
Agile analytics: nice and small.
Itility delivers agile analytics – with a team of data scientists we assist in exploring data with our Big Data SAAS solution (the “any tool” to upload your first data sets).
Back to overview