Where do I find logs?
Admin → Observability or your connected log stack (ELK, Datadog).
Logs include method, path, status, duration, user/org (non‑sensitive), and x‑correlation‑id.
Do you expose metrics?
Yes. Prometheus endpoint at:
GET /metrics
Scrape it and visualize with Grafana.
How do I correlate logs?
Use the x-correlation-id header we return on each response. Pass it in support tickets to trace a request end‑to‑end.
Can I export logs?
Yes. CSV/JSON export from the admin console or via API. Apply filters (time, actor, route, status) before exporting.
Which metrics matter most?
Error rate (4xx/5xx).
Latency (p95 request duration).
Payment success rate.
Dispute rate and time‑to‑resolution.
On‑time milestone completion.
Do you redact sensitive data?
Yes. Tokens, passwords, and confidential payload fields are redacted in logs. Attachments are never logged (only metadata).
Can I set alerts?
Yes. Use your monitoring stack to alert on high error rate, slow endpoints, missed webhooks, or dispute spikes. Set clear runbooks (next steps, owners).
What dashboards do you recommend?
API latency and error rate.
Payments funnel (intent → succeeded → refunded).
Work health (milestones on time vs late).
Disputes (opened vs resolved).
System status (SSE connections, webhook queue).