Skip to main content
The DataOps dashboard is the control tower for pipeline operations. You see which jobs are green, which are trending slower, and where data quality checks failed—all without opening each pipeline individually.

​
DataOps dashboard

Widgets typically include:
  • Run outcomes – success, failure, skipped over the last 24 hours / 7 days
  • Latency – schedule drift (did the 06:00 job start at 06:00?)
  • Throughput – rows or bytes processed relative to trailing averages
  • Ownership – teams responsible for the noisiest failures
Filter by environment (dev/stage/prod), domain tags, or connection to focus incident response.
Collapse to SLA tiles: on-time loads, critical dataset freshness, open incidents.

​
Health metrics

Health combines recency of success, error budget consumption, and dependency availability (for example, warehouse incidents). A pipeline can be “degraded” if it completes but misses freshness targets.
Pair health metrics with on-call rotations so degraded states page someone even when jobs technically “succeed” with partial data.

​
Quality scores

When you attach data quality nodes or contracts, the dashboard surfaces quality scores or failure counts per dataset. Drops often precede BI incidents; investigate before executives notice.

​
Schema monitoring

Schema drift indicators compare observed columns to expected schemas or contracts. New unexpected columns, missing required fields, or type widening/narrowing appear as events you can route to alerts.
Tune sensitivity so expected vendor experiments do not page nightly; whitelist known sandbox sources separately.
When multiple pipelines fail together, check shared infrastructure (warehouse, identity provider, DNS) before deep-diving each job.

Diagnostics

Drill into profiling and anomalies.

Alerts

Turn dashboard signals into notifications.