We updated Checkly AI (currently in beta) with new features based on user feedback in the last period. Here are the highlights:

AI Analysis for Error Groups

You can now trigger an AI analysis for any Browser Check or Playwright Check Suites check right from an Error Group. This works in the context of Test Sessions and regular Monitoring.

User Provided Context

You can now tweak the User Impact and Root Cause by providing your own context after the initial analysis is done and run the analysis again. Providing this context has many use cases:

• Ignoring errors that are expected or irrelevant.

• Pointing out specific errors that you know are critical.

• Adding knowledge about UX / UI patterns.

This is a first release. We will make user provided context a first class citizen in the future so you can document the quirks of your service and get relevant analyses out of the box.

Miscellaneous

• We now recognize graphQL errors correctly and provide them to our AI analysis.

• We now recognize various types of chunk loading errors correctly.

• We’ve move to a streaming, async solution to AI analysis in our web app is more performant and reliable.

Checkly AI is by default enabled for all users, except for Enterprise customers. If you’re an Enterprise customer, go to the Labs section and turn on “Checkly AI.” Then, opt in to the feature on the “AI” screen in account settings.