Tracehound

Tracehound reads errors from your monitoring tool, groups duplicates that differ only in IDs or timestamps, and ranks the recent deploys and commits most likely responsible. Each incident gets a short summary of what broke, how many users hit it and which code paths are involved.

On-call engineers at teams with frequent deploys use it to cut the first fifteen minutes of an incident down to one. It plugs into Sentry, Datadog and OpenTelemetry. The commit ranking is a probability, not a verdict; it is often right for regressions introduced by a single change and much less useful for failures caused by infrastructure or third-party outages.