Observability specialist

Iris AI

Stop jumping between disjointed tools. Iris manages Datadog, CloudWatch and Grafana from one interface.

A free 30-minute demo with an engineer. No commitment.

Iris AI, a typical day Under policy
  1. Iris AI

    p95 latency on checkout-svc up 38%. Correlated with RDS connection saturation. Paged on-call with the trace.

    Resolved 01:31
  2. Iris AI

    Proposal: exclusion filter for debug logs on indexer-*. 62% of indexed volume, 0 queries in 90 days. Archive to S3 with rehydration.

    Awaiting approval
  3. Iris AI

    14 monitors paged in the last 30 days with no action taken. Proposed: consolidate to 3 composite monitors on customer impact.

    Awaiting approval

Illustrative. Names and figures are representative of a managed account.

What it does

Built for Datadog cost.

Iris configures your monitoring across vendors, finds the metrics and logs you pay for but never query, and tunes alerts so a page means customer impact.

Unified multi-vendor configuration

Datadog, CloudWatch and Grafana managed from one place, with dashboards, monitors and log routing expressed as configuration rather than clicks.

Sensible setups, suggested

Recommended monitors and dashboards per service from what actually runs, so new services are covered on day one without copying last quarter's settings.

Alert noise removed

Monitors consolidated and thresholds tuned to customer impact. Alerts nobody acts on are deleted deliberately, with the reason recorded.

Observability cost under control

High-cardinality tags, unqueried metrics and fully indexed logs found and fixed: allow-lists, exclusion filters and archive routing with rehydration on demand.

How it works

From signal to approved change.

  1. Read

    Datadog usage by tag, index and host, plus 90 days of dashboard and monitor queries, through read-scoped keys.

  2. Find

    Metrics and logs nobody queries, monitors that page for nothing, and the tags that multiply your bill.

  3. Approve

    Each change proposed with what it removes, what it keeps and how to restore it. An engineer approves.

  4. Tune

    Monitors and routing adjusted continuously as services change, so coverage stays current and noise stays down.

Guardrails

What it will never do on its own.

  • Removals checked against 90 days of real queries first
  • Coverage is never reduced, only noise
  • Every log filter has an archive and a rehydration path
  • Changes land as configuration, with a record of what changed and why

Works with

Inside the tools you run.

DatadogCloudWatchGrafanaPrometheusKubernetesSlack

The other agents: Sage AI, Finly AI, ClearRisk.

Questions

About Iris AI.

Will Iris replace Datadog?

No. It configures the tools you already have and cuts what you pay for them. Consolidating vendors is possible but it is your decision, not a default.

How does it decide what to remove?

From usage: dashboards, monitors and searches over the last 90 days. Anything queried stays. Anything removed can be restored, and logs go to archive rather than away.

Does it handle on-call routing?

It routes pages with context to your rotation, which DevLift engineers can join. Escalation policies stay in your paging tool.

See Iris AI on your accounts.

Schedule a free demo with an engineer and watch it work on a live environment.

You'll pick a time on the next page.