Breaking Down Data Silos with the Fusion Artificial Intelligence Data Platform
All too often, I have found organisations collect and report their data in silos. Even when some data is collected in a single data model, report outputs still often remain siloed with no cross functional reporting.
Of course there will be times that siloed data is required, for example when related to a single functional process and within the Oracle Cloud ecosystem, there are tools dedicated for this purpose such as Financial Reporting Center for General Ledger data and Oracle Transactional Business Intelligence which has single-pillar data models known as Subject Areas to meet the needs of functional processors and managers.
But what about when users need to see data from across not only their Oracle Cloud platform, but across their entire enterprise? That is where the Fusion Artificial Intelligence Data Platform (FAIDP) comes in.
What is the Fusion Artificial Data Intelligence Platform?
The FAIDP has been known by several names since its inception. Starting off as Analytics for Applications or OAX, it has grown through phases becoming Fusion Analytics Warehouse (FAW), Fusion Data Intelligence (FDI[P]) to its current stature as FAIDP.
The FAIDP is a complete Analytics package for your enterprise, combining market leading infrastructure and data visualisation tools with pre-built, managed data pipelines for your Oracle Cloud data. This minimises technical debt as organisations do not have to build these from scratch and maintain them at every quarterly release of the source. In addition to Oracle Cloud data, there are native connectors to over 40 other data sources, enabling organisations to bring in data from any system they use to form a full Enterprise Data Platform.
In its latest iteration, it has expanded to incorporate the Artificial Intelligence Data Platform (AIDP) from Oracle Cloud Infrastructure, enabling customers to make use of advanced AI and ML techniques to help guide their organisations and make informed decisions and actions as efficiently as possible but that is for another blog.
How does the FAIDP help break data silos?
As noted above, data silos tend to form in two places:
Recommended by LinkedIn
On the storage side, the FAIDP breaks down the silos by streamlining enterprise architecture, replacing multiple data warehouses into a consolidated data lake. This in turn makes modelling data into a single enterprise model more straightforward, resulting in efficiencies through reduction in duplication of data and modelling effort across multiple solutions. Having all your data within a single platform also makes it far easier to model the data, combining attributes and measures from the different sources into a single data model.
Even once you have the model in place, you need to have your reporting requirements defined in a way that enables dashboard and workbook creation with a cross functional focus.
Often, I see organisations define their requirements focussing on business function e.g. what do I need to see to complete this task. This is where the siloed workbooks come from. Instead of looking from a transactional lens, it can create a more efficient user experience to define these requirements using a persona-based approach.
Take a head of department for example. Traditionally, to get an overview of how their department is performing, they would have to visit a general ledger dashboard to see how they are doing against budget, an invoice dashboard to make sure they are paying suppliers on time, a receivables dashboard to make sure they are getting paid efficiently before moving on to HCM, which may be in a completely different reporting tool to do the same there. With FAIDP, it is possible to combine these in a single place, regardless of the source, even combining these into joint metrics and bringing in front-office operational data. For example, the below workbook jas all the back office metrics needed in a single view.
There is a drawback with using the persona based approach. It may increase development times and costs in the short term as each business function will have multiple personas however this is likely to be more than offset longer term through improved user experience, better self-service adoption and operational efficiencies as users do not have to navigate around a complex catalog to find the individual reports they need across multiple dashboards, for example, Management Accountants and Budget Holders will both need to see the same finance reports for performance against budget but the management accountants will need to access other information such as purchase orders and invoices whereas Budget Holders would want to see whether increased operational activity has caused a variance. These would be two very different workbooks starting from the same place.
Key Takeaways
If you would like to discuss anything in this article or anything else related to reporting in Oracle Cloud, please contact me.
Great insights on breaking data silos, Daniel. In your experience, what’s the most persistent challenge when aligning stakeholders across departments to fully leverage FAIDP’s capabilities?