Checkly

Are you heading to KubeCon next week? ?? Join us with Rootly, DX and Coralogix for KubeCon After Dark, an after-hours spot for those who want more than another hotel lobby bar. ?? April 2, 2025 | 7:30 PM - 10:00 PM ?? The Lockdown Room, London No panels. No sales pitches. Just good drinks, better music, and the right people. ?? RSVP before spots fill up: https://lu.ma/twxvg313

?????? Forgot to post a reminder for "Playwright in Production" Webinar w/ Filip Hric tomorrow hosted by Checkly! We'll be: - data extraction for a focused knowledge base - maybe buying a car to demonstrate - walking through how you can use Playwright to validate and then continue to monitor environments you care about! Link of event in the comments Sorry Arnaud Fournier ??

Think the OpenTelemetry Collector ?? is just a data pipeline? Think again. It’s a powerful proxy, a flexible data processor, and a key tool for optimizing observability, without modifying your application code. But many teams don’t tap into its full potential. Join us for a deep dive where we’ll cover: ?? Common misconceptions about the Collector ?? Filtering and efficiency best practices ?? Statelessness and chaining multiple collectors ?? How Checkly uses the Collector to power Checkly Traces If you’re working with OpenTelemetry, this is a session you don’t want to miss ?? https://hubs.ly/Q03cTyJM0

?? Not too long ago, the Playwright team released 1.51, and it came with a new "copy as prompt" feature. Whenever your test cases fail, you can now press a button and copy a full-fledged prompt, including test details, a page snapshot, and the failed script... It's a massive prompt and far more comprehensive than what I usually provide to an LLM. ?? But does this mean that AI can now magically fix all the failing test cases? And should all tests be fixed in the first place? After playing around with it for a little bit, I discovered that LLMs shine when you're unfamiliar with the Playwright APIs or the given errors. But for gnarly failures, the fixes can sometimes be surprisingly good but also surprisingly bad. Providing as much test context as possible improves the AI fixes, but that's usually not enough. No AI in the world can know when your features work correctly. You always have to double-check if your "fix" is an actual fix. Don't let the LLMs just do their thing because then it's very likely that you'll end up with a passing test suite for a broken application. And that's missing the point, isn't it? ?? As always, you can find some more thoughts on YouTube. ??

How does an ambitious remote-first company with a growing team and trust as a core company value approach employee compensation? We recently spoke with Kaylie Boogaerts, Director of People at Checkly, to find out. Here’s a few of the most interesting pieces of their approach – and for the full details, head to the article linked in the comments below ??

We're hosting a livestream with Vercel on ?? April 1 at 9:00 AM PDT / 6:00 PM CET about Proactive Monitoring and Debugging. Simply follow the link below, join the community and click “Going”: https://hubs.ly/Q03cvzjF0 See you there!

Headed to KubeCon Europe and wondering where the after-hours is happening? We're hosting a high-energy, no-agenda party designed to keep the night going with Rootly, DX & Coralogix ?? April 2, 2025 | 7:30 PM - 10:00 PM ?? The Lockdown Room, London ?? RSVP now: https://lu.ma/twxvg313

Render powers millions of services across its cloud application platform — reliability is non-negotiable. But as the platform scaled, traditional monitoring wasn’t keeping up. ?? With Checkly, Render modernized their monitoring stack and reduced resolution time from up to a week to under an hour. ?? In addition, they also achieved results like: ? Automated full coverage across 10+ clusters with Terraform ? Implemented Checkly Traces via OpenTelemetry to pinpoint issues fast ? Started to proactively detect misconfigurations before they hit production ? Added collaboration-friendly workflows with Slack, tagging, and SSO Find the full story in the comments ??

I am absolutely convinced that I cannot expect any person on this planet to pay even a cent for our product, if I am not happy with it myself. This is why I dogfood and use Checkly every day for every imaginable use case. I want to know the strengths and weaknesses of our product so I know what we should be focussing on when we improve it. You can see that my team really appreciates me testing limits in Checkly ??

Playwright webinar incoming. Filip Hric and I are going to dive into three use cases; AI knowledge bases, automating tasks, and validation engineering ?? Link in comments!