Tokonoma profiles your services to decide what to monitor. With Agentic Monitoring enabled, it spots unusual behavior, groups related signals, and starts investigations using your team’s expertise and past findings.
Error rate rose above its recent baseline.
Request latency rose during the same interval.
Signal closure records a return toward baseline; it does not establish incident resolution.
Checkout connection timeouts
Repeated container restarts were recorded for checkout.
Signal closure records a return toward baseline; it does not establish incident resolution.
Rabbit message processing errors were recorded.
Signal closure records a return toward baseline; it does not establish incident resolution.
Only signals linked to retained detectors in this scope are shown.
When Agentic Monitoring is enabled, Tokonoma profiles your services and entities, discovers available telemetry, and proposes monitors. It checks those proposals against historical data to find useful signals before ongoing monitoring begins.
Related signals are grouped and compared with service context, signal history, and past investigations. With TypeSafe’s Jev enabled, Tokonoma uses that context to distinguish explained recurrences from findings that warrant investigation.
Review the affected service, supporting evidence, and related occurrences together. Tokonoma also identifies gaps in telemetry and alert coverage that can leave production problems unnoticed.