Datadog

Sources: Datadog (homepage) · LLM Observability

Stub

Observability / APM vendor with first-mover position on LLM and agent monitoring. Datadog AI Monitoring / LLM Observability cited at CMM D9 (Operations & Human Factors) for latency / cost spans on agent workloads, alongside New Relic AI Monitoring and Sentry AI Tracing. Built on Datadog’s existing distributed-tracing pipeline; integrates OpenTelemetry `gen_ai.*` semantic conventions.

Pending content: company overview, full LLM-observability product description, integration with agent-aware SIEM playbooks at D7, public case studies.

At Unprompted (March 2026)

Datadog presented two talks at the Unprompted Conference (March 2026). In Exploring the AI Automation Boundary for Threat Hunting, Arthi Nagarajan describes the migration from a single threat-hunting agent to an orchestrator-subagent system and frames the “automation boundary” — the line between where AI accelerates defensive work and where it introduces new risk. A second talk, by Olivia Gallucci, covers AI-assisted macOS vulnerability research.