agent-observability-session-classify

You can classify user intent satisfaction in Datadog Agent Observability traces by analyzing session interactions and audit trail data. This Skill evaluates conversation content and RUM behavioral signals to derive a satisfaction verdict.

Yes, RUM data confirms session classification by providing behavioral signals. The Skill utilizes Real User Monitoring data alongside audit trail analysis to validate whether a user's intent was met in an AI agent session.

You can evaluate AI agent effectiveness by classifying sessions using session_id, trace_id, or ml_app modes. The Skill retrieves conversation content and metadata from these specific modes to determine interaction success.

Detecting user intent satisfaction requires Datadog Agent Observability, RUM, and audit trail data. These sources provide the conversation content, behavioral signals, and server-confirmed effects needed for classification.

Yes, classifying Datadog Agent Observability sessions requires access to Datadog Agent Observability and RUM data. The Skill needs these permissions to retrieve and analyze conversation content and behavioral signals.

Analyzing the audit trail classifies AI agent session satisfaction by checking for server-confirmed effects of the assistant's actions. This confirms whether the agent's automated responses successfully satisfied the user's original intent.