ANNOUNCEMENT:
A year ago, when we launched MultiTool, we followed the traditional wisdom of launching an MVP. We found an unsolved problem in DevOps with a real market need, built a solution, and launched it as quickly as possible. But sometimes "minimum" isn't good enough.
Our design partners loved MultiTool's capability, but the DX was rough. For the typical engineer, the deployment process has a ton of inertia because deployments are so load-bearing. If something goes wrong with your deployment process, the whole team suffers, so it's hard to find an internal champion willing to take on the political risk necessary for buy-in. If the tool isn't perfect, it could be painful. And perfect means mature.
In short, our MVP did not deliver. That's why we rethought MultiTool from the ground up.
For the past five months, I've been quietly rebuilding MultiTool in totality for our upcoming relaunch. Now, MultiTool is focused on monitoring and alerting. Unlike v1, it's a zero-config, zero installation process. Point your OTel data at our server, and get automatic 24/7/365 monitoring and alerting.
Our focus is shifting from automated canary analysis, which will no longer be supported on the new platform, to combating alert fatigue. The same thresholdless metric analysis algorithm that powered the canary analysis engine now powers the new monitoring platform.
Alert fatigue is a problem only made worse by the adoption of AI, as there are more services, deploying more often, with more bugs, requiring more hand-tuning. With MultiTool, there is no hand-tuning. We model your application's baseline performance and alert you whenever there's a regression, pinning down the commit and diff that caused it. Automatic root cause analysis with no config, on any codebase.
I'm so excited to launch what we've built, and while all the features won't come out at once, we're moving faster and building more than ever before, and the future of Wack Incorporated shines brightly through the polychomatic glass pane of innovation.
