Imagine that in the apartment above you, your neighbor turned on all the faucets, plugged all the drains and water started leaking into your home from above. You need to catch all the water to keep it from ruining your house, but water is leaking from multiple places at different rates and you only have one bucket.
Think of that water as your data. Even with a really big bucket or even a drainage system to process all the incoming water, you can’t collect and manage all the water with one bucket. This is why we have, and will always have, data silos.
It would be convenient if one bucket could catch all the water, but you need to catch water coming down in different rooms at different rates. Okay – I don’t want to drown you in this analogy (but see what I did there?).
The point is that we need to stop talking about “breaking down data silos” in the context of enabling analytics and instead, think about the approach to breaking down insight silos, which are the inconvenient offspring of data silos.
We must first accept that that data silos are not pure evil. They are necessities.
- Some data silos help us ensure better security, by creating physical separations that help us control access from apps, systems and people.
- Some data silos collect streaming data from IoT devices on the edge and enable real time AI.
- And some data silos provide access for business analysts and data scientists to experiment with data while being sure to not disrupt critical IT services that tap into other silos.
Instead, data analytics stakeholders need to think differently and focus on the insights, which are no longer physically bound to the data itself – no longer constrained by the walls of the silo – free to grow, make friends with insights from other silos, and blossom into more insightful insights! Yes, modern technology has come to the rescue and unshackled these hidden insights from their data silo guardians, regardless of how well they’d been hidden. Although data lakes, data marts, data warehouses, etc. help keep data silos off the streets, there will always be many data silos that will need to be overcome.
Hence, breaking down the data silo is not the concern of the data analytics stakeholder – it’s about looking across the silos and breaking down the insight silos. An analytics platform that is architected to access, collect and process all types of data (structured, unstructured, internal, external, real time, etc.) for analysis can unify the insights that are hiding in data silos. For example, TCS Connected Intelligence Platform (CIP) is a scalable enterprise insights platform that drives automated intelligence into your business. CIP unifies big data management, real time insights, stream processing and AI/ML engines and fulfills customers’ end-to-end needs for easily building analytic use cases from all types of data – from data ingestion, to data lakes, data wrangling and advanced analytic modeling.
Importantly, a unified approach like this serves to reduce time to value and TCO with accelerated insights that span all types of data. Both business and data practitioners stand to benefit as they learn to power their business models with these systems of insight – embrace data silos while avoiding insight silos.