Hero Image

The Single View Challenges: 1 Big Data


29 Jan 2021 | Emma Benham

In a demanding world, large teams with distributed business units and disparate systems are hard to consolidate. As organisations find themselves with legacy products, increasing pools of data and knowledge silos the need to aggregate this information to generate value is real. However, many have struggled to harness the power of Single View and unlock the true potential of their data. In this blog series gravity9 break down the six fundamental challenges organisations face when setting seeking to achieve Single View and how through uniting deep technical knowledge with the art of senior problem solvers Single View doesn’t have to be an aspiration, it can be a reality.

Challenge 1. Big Data

Big data is a good problem to have. It’s is also the most common and challenging problem organisations face when seeking to achieve a Single View solution.

Let’s breakdown the problem. The benefit of greater data pools is that it provides deeper insight and offers a greater opportunity to evaluate opportunities, based on historical facts. Essentially, it is the equivalent to human experience for a complex AI algorithm. However, drawing your data from disparate, siloed systems is not only a logical challenge but, a physical one too. Responding to this recent realisation of data requires a modern approach to storage that is able to manage and process the data quickly.

So how do you solve this solution and create a Single View?

Operational data stores. Operational aggregate stores are often compared with more traditional data warehouses, however, the aggregated operational data store has very different demands for accessing and reading cross-system data. Warehouse solutions manage the vast amount of data through de-normalisation which is reasonable for the purpose of reporting and even business intelligence, however, the Single View requires a lower level. There is a need to see transactions, historical detail and facts to make the solution efficient and detailed for operational users and algorithms.

When creating solutions, we often partner with MongoDB who provide a database designed specifically for large scale variable data with inbuilt flexibility to store data without a predefined schema. For this solution, data is stored as objects (aggregate roots) complimenting effective caching strategies. The scalability comes through the ability to shard information; the act of spreading data and process loads across multiple physical nodes. This enables the data to be centralised logically but distributed physically. Sharding is a powerful concept which enables the storage of information from many systems at a granular level and the ability to access it at speed, something often difficult to achieve in more traditional relation technologies.

Is your data just rolled up or true single View?

Aggregating broad data sets can result in ‘rolled -up’ summaries of the data, which in turn produce a low level of data granularity. Within Single View, all data is mapped globally at a transactional level ensuring users can view and engage with the data at a high level as well as the low level (for example: seeing individual transactions based on a single customer or service). This is achieved using data warehousing solutions that roll-up the data for easy consumption while maintaining complete data integrity.