Datadog Integration
Koalr's Datadog integration reads deployment markers, APM error rates, and SLO burn rates to complete the deploy risk feedback loop. When a deployment causes an error rate spike or SLO burn, Koalr labels it as a failure — automatically retraining the risk model for your organization.
What Koalr pulls from Datadog
- Deployment markers — custom events tagged to production deployments
- APM error rate and latency percentiles (P50/P95/P99) before and after each deploy
- Service health metrics — error rate delta within 10 minutes of deployment
- SLO burn rate at deployment time — real-time budget consumption
- Infrastructure metrics — CPU, memory, and request throughput per service
- Log events correlated to deployment windows
How Datadog data feeds into Koalr
- Deploy risk outcome labeling — error rate spikes confirm or refute risk predictions
- SLO burn rate signal — high burn rate at deploy time elevates risk score
- MTTR measurement — incident duration from Datadog monitor alerts to resolution
- Change failure rate — service health degradation within 10 minutes of a deploy
- ML model feedback loop — Datadog confirms outcomes that retrain per-org signal weights
How to connect Datadog
- 1
Go to Settings → Integrations in your Koalr dashboard.
- 2
Find Datadog in the integrations list and click Connect.
- 3
Generate a read-only API token in Datadog and paste it into Koalr. No OAuth flow required — the token is encrypted at rest.
- 4
Koalr begins syncing historical data immediately. Most integrations backfill 90 days of history on first connect.
Permissions and scopes requested
Koalr requests the minimum permissions required to read the data above. All access is read-only unless noted otherwise.
DD-API-KEY (Datadog API key for metrics and events access)DD-APPLICATION-KEY (application key for SLO and dashboard data)
Ready to connect Datadog?
Connect in under 5 minutes. Koalr backfills 90 days of history automatically — no manual imports, no CSV uploads.