Working for a company that has been delivering live Big Data projects in banks since 2010, it is interesting to see the latest thoughts around the best strategies to run Big Data projects.
It seems that the general strategies of running with small teams with a cross between, the Business, Architecture/ IT staff, Data Scientists - in other words a small cross practice team to deliver solutions in short time frames using Agile methodologies as part of an Insight Approach, is exactly what we do as a business.
We have found that the methodology as outlined by leading IT Research companies (not mentioned by name) is the methodology we have adopted over the years realising that boiling the ocean approaches, are (relatively) costly, do not work and people end up with large lakes full of data they cannot really cannot access.
The key to the Strategy is: Operate through the actions of agile, cross-functional insight teams.
Close the loop between action, insight, and data for a specific, priority outcome
Collect big data as part of the insight-to-execution process. As a company we take a pragmatic approach, delivering use cases, which then lead to delivering value on an iterative approach, allowing added-value to be accessed as soon as possible.
Many approaches allow companies to deliver within a couple of months or so, achieving what was only dreamt of in the past at low cost.
Our expertise is growing across sectors, covering Banking and Insurance (Financial Services), Energy and Utilities, including Data Providers and moving to other sectors such as Retail.
By leveraging behavior-driven insights, systems of insight apply advanced analytics to operational environments at the point where they generate customer interaction data for immediate deployment via systems of engagement. How does a system of insight work? The building blocks, or model by which a system of insights functions is comprised of the following capabilities: Consume – Ingest structure, unstructured, and transactional data Collect of data Report – Understand what happened Detect – Identify patterns and correlations Decide – Implement logic associated with business rules