The general idea of data democratization and providing access to everyone in the company to use it is a very simple idea, the realization of this idea is however a very complex one in many cases. This is especially true in organically grown companies who have, over time, grown their IT footprint. In general, this includes a large set of legacy applications who do not by nature support integration that well.
However, the fact that an enterprise has a large set of legacy applications should not hold back the ambition to change to a more data driven enterprise. Moving to a more data driven enterprise, democratization of data and base decisions on actual data is a huge benefit for enterprise. Additionally, it is the starting point of integrating other systems and drive business in new and disruptive ways to keep the advantage over competitors.
To get started with data democratization the first step is to start finding your data and classify the data sources. The below pointers can be of importance when evaluating the data.
- Data location : where is the data located, how easily can it be accessed
- Data ownership : which department owns the data
- Data confidentiality : How confidential is this data
- Data privacy : is there privacy related data in the set
- Data value : what is the monetary value of the data
- Data alignment : how well aligned is the data with other sources
Taking the above questions into mind when classifying all data this will give you a route of action per dataset. I will help you to identify how to handle each data-source, how to classify it and to integrate it. It also helps you to prioritize it.
Moving to the cloud
When moving to a data democratization model, this might be a turning point in how you look at IT and it might be good moment to consider the use of cloud. When trying to integrate and store a large set of data you can select, as an example, the Oracle Cloud to house the data you make available for all your users.
This is not necessarily meaning that you have to move the actual systems to the Oracle Cloud. One can think of a model where the backend systems remain in your current datacenter or cloud and you move / sync your data and the changes to the Oracle Cloud where you unlock them to the users using REST API’s and portals in the form of a data shop.
Opening up with a data shop
The concept of a data shop is the way to get started with data democratization. A data shop is a self-service portal where users can gain access to all the data that you have liberated. It provides users the option to get access to REST API’s or to, as an example, the Oracle Data Visualization Cloud Service, which can show data already included in graphs and other visualization.
As with a real shop, a large number of “products” are available. Some are for the standard users in the form of pre-defined dashboards and reports and some users will require the data in a rawer format to make and share their om reports and analysis.
Making it easy
Making it easy for data consumers to use the data is actually two folded. You will have two types of consumers, the tech consumers and the non-tech consumers. The tech consumers will require REST API’s to gain access to the data and undertake all the actions they need and think are valuable. The other type of consumers are the non-tech users. For non-tech users the REST API approach might be too difficult to master and they will need a more simple way to gain access to the liberated data.
After you moved your data to the cloud , as a first step in the process you will have to ensure that the data is accessible, via REST API’s and also via standard dashboards. Oracle is providing a growing number of options in the Oracle Public Cloud to do both. You can use standard visualization and data exploration tooling to your users within the cloud which have a relative low learning curve and people can start with right away. An example of this is the Oracle Data Visualization Cloud Service.
Oracle is also providing API functionality, even though the services Oracle provide standard from within the database and with some of the cloud services it might very well be beneficial to consider building your own REST API implementation while leveraging both the Oracle Compute Cloud Service with Oracle Linux instances and the Oracle Container Cloud Service.
Putting it all together
With data democratization, you open up your data, break the silo way of architecture and provide your users the option to analyze the data and make use of a active and up-to-date collection of data in one single place, the data shop. Moving to the cloud and leveraging the cloud is a technical solution to make this happen. Moving to the cloud is not the goal for data democratization.