Troubleshooting a troubleshooting platform
Product Forward worked with a Splunk Partner to redesign an internal tool that cut hours off task completion
Project Overview
The challenge at hand was the time it took Splunk technicians, using an internal diagnostic tool, to resolve customers’ issues — it was too long with too much room for error. A Splunk engineering partner to approach to us to help them figure out how to make them faster. We worked together to come up with a solution to troubleshoot the platform that troubleshoots problems. What a unique opporunity!
Project Execution
The engineering partner got to work and completed a proof of concept that could detect problems based on diagnostic files the customer was able to capture using the new tool. This was presented to Splunk leaders who loved it. To bring the idea to life, they looked for a partner who could bring experience to life and deliver results.
Our first approach was to find out what Splunk users need and how they behave, and integrate that into the solution. We shadowed the customer support team andbegan learning how customer support does its job around the Splunk product, we sought not to understand the complexity of the product, but rather to make sense of the behavior of the people using it.
After a series of collaborative sessions with programmers, support technicians and service implementation Splunk experts, solutions emerged as we progressed through rapid prototyping cycles.
Results
The final product was test with a control group of untrained Splunk support technicians, who received the beta version of Splunkbot. Training typically takes 8-12 weeks but within just a matter of hours, these new support technicians were able to troubleshoot customer cases using Splunkbot, solving cases at a fraction of the time.
