Some problems are characteristic of a specific platform, and some platforms fail way too often, deteriorating the CI experience for all the other platforms.

Create a dashboard that

• Selects a specific type of task (mostly HealthChecks as they should always pass)
• Finds all relevant workitems for such tasks within a time frame
• Counts the number of failed/error integrations, grouped by {workitem_action, config_identifier}
• Shows the config-identifiers that fail most often.

Another dashboard that shows the most unstable tests for a selected configuration.