Datadog Integration
Ingest Datadog incidents and tagged deploy events for Quality and DORA.
Why connect Datadog
Datadog is a Monitoring & Observability source for source ingest only. After you connect, AI Driven SDLC reads your org’s incidents and deploy / change events (when they carry commit or environment tags) on a schedule and uses them for Quality (incident rates, MTTR) and DORA (deployment frequency) views.
The same connect path works for every customer workspace. Platform self-observability (RUM, APM, or log shipping into Datadog) is separate and is not configured here.
The integration is read-only. We never create or change monitors, dashboards, or incidents in your Datadog org.
What you need
- A Datadog organization on one of the supported sites below.
- An API key and an Application key with read access to incidents (and change/deploy events if you want deploy metrics).
- The site that matches where your Datadog account lives (wrong site fails validation even with good keys).
| UI label | Site value |
|---|---|
| US1 | datadoghq.com |
| US3 | us3.datadoghq.com |
| US5 | us5.datadoghq.com |
| EU | datadoghq.eu |
| AP1 | ap1.datadoghq.com |
Gov cloud and custom API hosts are not supported in this release.
Select Datadog in your tech stack under Monitoring & Observability if you use it today.
Setup
- In Datadog, create an API key and an Application key with incident read (and change/event read if you care about deploys). Prefer the least privilege that still lets validation and incident polling succeed.
- In the platform, go to Settings → Integrations (or onboarding Integrations) and choose Connect on Datadog.
- Enter API key, Application key, and site.
- Click Test connection. Connect stays disabled until the test passes. A successful test confirms the keys against Datadog’s validate API for the chosen site and may show a short org summary.
- Click Connect (or Save) to store encrypted credentials. Keys are never shown again after save (the UI only reports that each key is present).
If the test fails with invalid credentials, rotate or re-copy both keys. If keys look right but the test still fails, switch the site (US1 / US3 / US5 / EU / AP1) to match your Datadog account region.
An empty org (no incidents or tagged deploys in the lookback) can still connect and test successfully. The connected card then shows 0 incidents and 0 deploys for the last 30 days.
Workspace viewers see the Datadog card as read-only; connect, disconnect, and sync need a role that can edit integrations.
Data we read
| Datadog data | Platform use |
|---|---|
| Incidents (open → declared; closed → resolved) | Quality metrics, MTTR; SDLC Insights “Time to Restore Service” |
| Deploy / change events with a commit SHA and/or environment tag | DORA deployment frequency; SDLC Insights “Deployment Frequency” |
We fetch roughly the last 30 days on each sync (scheduled daily or when you click Sync now). Deploy events without commit or environment tags are skipped; they do not count as failures.
Data we do not read or change
| Not accessed | Notes |
|---|---|
| RUM, APM traces, or log bodies | Customer source ingest only; no platform telemetry setup here |
| Notebooks, dashboards, or monitors as editable config | Not managed or embedded |
| Multi-org picker beyond one site + org per workspace | One Datadog source per workspace in v1 |
| Writes to incidents or events | Read-only |
Repository Readiness observability scans (dimension E) remain repo-based; connecting Datadog improves live Quality/DORA metrics, not automatic readiness score bumps.
Sync schedule
After connect, the platform syncs on a daily batch schedule. Use Sync now in Integrations → Datadog for an immediate pull. The connected card shows site, last sync time, and 30-day incident and deploy counts. You receive an in-app notification when a sync completes or fails.
SDLC Insights
Under Insights → SDLC Insights, Datadog sync contributes to the same weekly dashboard buckets as other monitoring sources: MTTR from incidents and deployment frequency from tagged deploy events.
If a tile shows unavailable or zeros unexpectedly, confirm Datadog has incidents or tagged deploys in the last ~30 days and run Sync now.
Disconnect
Remove Datadog under Integrations → Disconnect. The source is removed, further sync reports not connected, and polling stops. Previously ingested events remain in historical metrics until retention policies apply.