Splunk Capacity Models

While at Capital One, I colloborated with teams to create real-time big data dashboards using a machine-generated big data tool.

My last team at Capital One was responsible for capacity modeling throughout the Retail Bank line of business. After joining the team, I was put in charge of these capacity models. Before assuming responsibility, capacity models were done in Excel on a monthly basis with limited data points. Once I became familiar with Splunk, I saw an opportunity to automate this process. While Splunk was not designed with data modeling in mind, I was able to modify different functionalities within the tool so that I could automate capacity models throughout the Retail Bank LOB. The Splunk capacity models allowed leadership and platform teams to see how much traffic their platform was getting, how much of that traffic was from the web or mobile application, how much traffic a specific platform could handle, forecasts of future traffic, and more. Over the course of the year I was working on these capacity models, I increased the number of platforms that had capacity modeling from 2 to 8.

The new and improved capacity models had thousands of data points that provided more accurate data and projections. The revised capacity models could also be run ad-hoc versus once a month. Unlike the Excel capacity models that took half a day to a day to complete, the new Splunk capacity models ran automatically in a matter of seconds.