APM helps customers troubleshoot issues. Leveraging ML integrations to surface some of these issues can help our customers cut down any downtimes they might experience if we catch these issues and surface root cause before they have a wider impact.

Summary of the problem
I want to see any new/unique errors that are being thrown by my application. These unique errors could potentially be related to faulty deployments and could impact a larger subset of my customer base.

I want to view if a particular user agent is resulting in more errors than the others. E.g. I'd like to know if the customers using my webapp are seeing more errors if they are running a version of Firefox or Chrome.

Ideal solution (optional)

List known (technical) restrictions and requirements

• One click integration to start the ML jobs to start profiling errors.
• Bubble up the unique errors to distinctly identify these errors in the APM UI along with all the metadata associated with these errors.

Are there any pages or actions that relate to this feature?
For example: link to other solutions for further investigation