Spot missing rows before your business does

Monitor every pipeline, catch issues early, and keep your data trustworthy.

Failed runs reach your team first
Alerts are sent to workspace owners or the entire team via email and the Slack channels you already monitor.
- Email alerts
- Slack alerts
Alerts link straight to the logs
Each alert links to the run and its logs, so your team can investigate the cause before stakeholders notice.
- Runs, tables and logs
- Alerts with agent insights
Row counts show the data arrived
Rows loaded for each run show whether yesterday’s data is complete, even when every run succeeds.
- Rows loaded per run
- Data quality checks

A successful run doesn’t mean the data was received

A pipeline run can finish with no errors yet still load fewer rows than usual. The run status tells you the code finished. The row counts tell you the data arrived.

Pipelines Jobs list: my_pipeline's Rows loaded sparkline, with a tooltip showing a run that loaded 100 rows and succeeded — Rows loaded per pipeline
Docs
Rows loaded appear next to each pipeline’s run status, making short loads easy to spot at a glance.

Pipeline Runs for orders_api: every run succeeded, but the latest loaded 100 rows against 935 on the runs before it

Observability comes first when agents build more of your pipelines

An agent’s change can pass review and still change what a pipeline loads. dltHub shows the row counts of every run. Your engineers use them to check results before giving agents more work.

See background agents Explore the AI harness

Detect observable and silent failures

The run fails

dltHub sends your team an alert via email and Slack, with a link to the run.

Expired credentials
A crashed container
A timeout or memory limit

The load is incomplete

dltHub shows the rows loaded for each run, by pipeline and table.

A source that is not ready in time
An incorrect pagination field
An empty load

Find an incomplete load before it reaches a report

Monitor pipeline health, validate successful runs, and inspect detailed logs to understand pipeline behavior and diagnose potential issues. Docs

Walkthrough of the orders_api pipeline: runs that loaded 100 rows instead of 935, the run details, the job run, and the log lines where the paginator stops after one page — Pipeline run details and logs
Trace a successful run from status to logs, confirming expected data and completion.

Your team knows immediately when a run fails

Each email and Slack alert links directly to the run. dltHub also sends alerts for failures that your code cannot report, such as container crashes and timeouts.