oracle_hyperionThe term “performance” can mean different things within a given corporate context. In financial firms it most often refers to profit or loss. In tech companies, it can refer to throughput of an application or service. In this case, the term “performance” shares both of those connotations. The company was upgrading their Oracle Hyperion Financial Management installation, which is used to measure the company’s financial performance (and supports other tasks, such as forecasting). And because of the global nature of the company, and reliance on detailed ledger data ETL’d in from remote manufacturing locations, there was also a system-level performance aspect to the assignment. Due of the critical nature of close period reporting, there was a high degree of sensitivity around timely submission of data. This entry in my portfolio will focus on that latter portion of the “performance” equation, though the implications to the former will become clear as this entry progresses.

Beginning With The End In Mind

The above visualization is one of several artifacts that stemmed from collecting and collating data out of a combined Hyperion/Accelatis deployment. It provides snapshot views of requests that originate from company offices around the globe. Locations, request types and time interval can all be tailored to suit the viewer’s interest. It is used as a component for production support teams to actively monitor consolidations and other close period activities during critical windows of operation.

There are other visualizations based on this schema which was built in PowerBI, and are demonstrated through this portfolio entry. Another useful “end point” to consider is the schema diagram shown here. As I will detail later in this portfolio entry, the Concurrency table is the hub, as it has a row for every “grain” in the temporal span. From that, all time-based tables have a field with many-to-one relationship mapped to that central table. As visualizations are presented, the nature of how that hub maintains consistent time-scoping between various elements of a given chart will become apparent.

In the following pages, aside from the technical walkthrough, I will also detail some of the business and justifications for the process as a whole. While it would certainly help to have familiarity with Hyperion’s lexicon, it’s certainly not a requirement. This should not be considered a primer on Accelatis or Hyperion. Terms from those applications will be used in passing, with focus on how the data is collected and visualized. My goal is show how meaningful business process information can be gleaned from low-level data that the vendors themselves have not recognized. So, I’ve made an effort to keep the language as broad as possible.

keyboard_arrow_up^