Tableau and Looker are great for business dashboards. They're overkill for monitoring your own pipelines and scheduled jobs. QueryFlow's Observatory dashboard is native Mac, included with the subscription, and built for the data engineer.
Quick answer: QueryFlow's Observatory dashboard is a native macOS interface for monitoring your data pipelines, scheduled jobs, and query history. It shows success rates, average durations, error context, run history, and one-click retry for failed runs. Included with the $299.99/year QueryFlow subscription. Built for engineer-self-monitoring, not for sharing with executives.
Business dashboards (Tableau, Looker, Sigma): for executives and analysts, focused on KPIs and trends, expensive enterprise tools. Application dashboards (Datadog, Grafana): for engineering operations, focused on system metrics, paid SaaS. Pipeline monitoring dashboards: for data engineers monitoring their own ETL infrastructure, traditionally rolled by hand or built into the ETL tool. QueryFlow's Observatory is the third type, included natively.
Scheduled jobs view: every scheduled pipeline or query with next-run time, last-run time, success rate (last 30 runs), average duration. Recent runs view: timeline of all executions across all pipelines, filterable by status (success, failed, running), sortable by time or duration. Pipeline detail view: full execution log for a specific pipeline run, including row counts, error messages, retry attempts. Query history view: every executed query (manual and scheduled) with full SQL, execution time, row count.
When a pipeline fails, Observatory doesn't just say 'failed' — it shows the actual error message from the source or destination system, the source row data that triggered the error (when applicable), the SQL or query that was running, the timestamp and run-N attempt counter, the previous-run comparison (was this consistently failing or just this run). All the context you need to diagnose in one place.
For most failure types, one-click retry from Observatory is the fix: transient network issue, temporarily-locked Snowflake table, Salesforce rate limit. For persistent failures, Observatory's history lets you see when the failure started, which tells you whether it's a recent regression or always-broken state. Custom retry policies per pipeline let you configure max attempts and backoff for different failure tolerances.
Datadog: $15+/host/month for the basic tier, comprehensive but expensive for data-pipeline-only monitoring. PagerDuty: ~$25/user/month, alerting-focused. Custom monitoring with Grafana + Prometheus: free but requires self-hosting. QueryFlow Observatory: included with the $299/year subscription, focused on data pipeline monitoring specifically, no separate infrastructure to run.
Observatory is for self-monitoring (the engineer who built the pipelines monitoring their own work). For team-wide visibility into pipeline health across many engineers, cloud-based monitoring (Datadog, Honeycomb) makes more sense. For executive-facing pipeline health summaries, scheduling daily pipeline-status reports to Slack via QueryFlow itself (using the webhook destination) is the typical pattern.
Observatory runs as part of the QueryFlow app, not a separate process. Its data (run history, schedules) lives in QueryFlow's local SwiftData store. No noticeable memory or CPU overhead beyond the app itself. The dashboard renders results instantly even with thousands of historical runs.
v1.5 Observatory data is local to each QueryFlow installation. Cross-machine sync (via iCloud) is on the roadmap. For now, each Mac running QueryFlow has its own Observatory showing that Mac's pipeline history.
Yes. The query history view captures every executed query in the SQL editor, not just scheduled work. Filter by source (manual vs scheduled) if you want only one type.
Yes via the webhook destination. Configure a pipeline that watches Observatory's failure state and posts to a Slack channel or sends an email. The setup is more explicit than dedicated alerting tools, but more flexible.
Default retention is 90 days for detailed run logs and 1 year for summary statistics. Configurable in Settings. For longer retention, export the data to a warehouse via a pipeline (yes, you can use QueryFlow to monitor QueryFlow).
Yes. Observatory on a Mac mini shows the pipelines running on that Mac mini. If you want a unified view across multiple Macs, that's the cross-machine sync feature on the roadmap.
14-day free trial. Build a few scheduled pipelines and watch Observatory fill up with run history.