Better dashboards, better context, faster root cause

When incidents hit, responders don’t want more data, they want clarity. Over the past few weeks, we’ve delivered several improvements that help Aptible AI users move from detection to understanding faster:

Dashboard upgrades

• Diagnostic Dashboards now surface LLM-generated summaries of observed anomalies to provide a head start on root cause hypotheses.
  
  

  

  
  
  

• Dashboards display a live dependency graph, highlighting both impacted resources and upstream/downstream connections.

Observability infrastructure

• Users can now retrieve and render aggregated service-level metrics, making it especially easy to handle container churn.

• Historical metrics now persist cleanly even when containers are cycled or deleted.

• Observation permissions are now updated to account for deleted resources.

These changes not only make dashboards more useful, they also lay the groundwork for deeper integration with our knowledge graph. On-call engineers can now see past anomalies and relationship edges across dashboards, improving situational awareness with every alert.