What can an AI agent do with Datadog?

A Datadog AI agent can list and read monitors across your account, query time-series metrics, search logs, fetch infrastructure events, browse dashboards, and mute monitors during maintenance windows. It can generate health digests, correlate alerts with deploys, and surface what's actually worth paging someone about. Write actions require your approval.

Which Datadog data can my agent access?

Your agent accesses whatever is visible to the connected Datadog API key. You can scope tasks to specific services, environments, or monitor tags in the task instructions.

Can my agent mute or modify monitors automatically?

Yes, but muting and other write actions require your approval in safe and plan permission modes. You can configure auto-approval for trusted recurring workflows like scheduled maintenance windows.

How is a Datadog AI agent different from Datadog alerts?

Datadog alerts fire when thresholds are crossed but can't reason about what they mean together. A Datadog AI agent reads across monitors, metrics, and logs to write a plain-language summary of what's degraded, why it's happening, and which issues are worth acting on first.

Developer tools

Datadog, run by an agent.

An AI agent that monitors your Datadog metrics, alerts, and infrastructure health.

Connect Datadog Book a demo

What needs your approval

Datadog never changes anything without you.

Anything that creates, edits or deletes in Datadog waits for your approval. Approve a recurring task once and it runs unattended after that.

Connecting Datadog

About 5 minutes

Auth: API key
Setup: About 5 minutes

Where to get access

Create an API key and Application key in Datadog under Organization Settings. Enter them as API_KEY:APP_KEY.

app.datadoghq.com

Example tasks

Real work people hand to an agent that uses Datadog.

Generate a daily infrastructure health digest and post it to Slack with any firing alerts
Monitor p99 latency and error rate after each deploy and summarize regressions
Detect a spike in errors across services and correlate with recent events in the stream
Produce a weekly on-call report: alerts fired, time to resolution, noisiest monitors