Automated ticket triaging

Netflare was founded with the goal of providing a comprehensive system for managing and responding to incidents and crises. As we developed our MVP, it became clear that we needed a ticketing system to help our customers keep track of their issues.

As an engineer who had spent a lot of time being oncall at Amazon, I thought I was well-prepared to build this tool. But as we developed the MVP, I realized that something was missing. I looked at competitors, but didn't find anything that truly inspired me. Most ticketing systems are simply places to document and track issues until they are resolved, and I wanted to create something more.

It finally dawned on me that the problem was the fact that I was spending so much time in the ticketing system. The incident itself can't be solved within the ticketing system, yet I was spending a lot of time there because nothing was being automated for me. When you have 50+ open issues and more coming in every hour, it takes a lot of time to triage them, which takes away from actually solving the problems. Most ticketing systems are built to leverage computers to store your issues, but haven't innovated beyond that.

Once we realized this, we had a clear vision for bringing ticketing into the 21st century by automating the triaging of issues so that users can focus on the problems themselves. This required a major overhaul of both the User Interface and User Experience. With the help of talented UX designers, we were able to come up with several solutions.

First, we needed to reconsider how you go through your list of issues. We found that the traditional way of giving you a list of all open issues with links to the ticket detail was a time sink as you jumped between ticket to ticket to get caught up. We therefore re-designed the list so that you can peek into it without having to jump between pages so you can quickly get caught up. Second, we made the list more like an email inbox, indicating which tickets are new or have updates, so users can easily identify the issues that need their attention. Third, we added an announcement feature that allows users to easily keep stakeholders up-to-date on the latest developments. And finally, we enhanced our pager integration so that users can set up workflows to be notified of issues they care about as soon as they happen.

I'm proud of what we've accomplished so far and am excited to announce the re-launch of our ticketing system, which we now consider to be in the "Minimum Loveable Product" stage. We are still learning from our users and are eager to implement more automation to improve the oncall experience.