Dashboard

Monitor Import Tool health, run outcomes, and stream freshness from the Dashboard. Set thresholds that determine whether streams are healthy or unhealthy.

The Dashboard is the first view when you open the Import Tool perspective. It summarizes import volume, success and failure rates, stream health, and activity over a selected time range so you can spot issues without opening each import.

Import Tool domain tools are available when you create or edit an AI agent. See AI Agents.

Dashboard with job and stream filters, KPI cards, and import charts

Filters and layout

At the top of the Dashboard you can narrow the view and share or refresh it:

• Job group – filter by a job group. Create and maintain job groups in Manage groups....
• Jobs and streams – restrict the metrics to selected imports or streams, or leave All jobs and All streams.
• Time range – choose 24h, 7d, 30d, 90d, or Custom.
• Frequency – choose how the charts group data in the selected range: Daily, Weekly, or Monthly.
• Refresh – reload your import metrics.
• Copy link – copy a link to the current filtered view.
• Edit layout – rearrange the dashboard widgets to suit your needs.

KPI cards

The summary cards show tenant-wide results for the selected filters and time range:

Card	What it shows
Entities imported	Total imported entities, with a breakdown of new versus updated records
Success rate	Percentage of successful runs in the selected period
Failed records	Records that failed, with the number of failed runs
Health	Overall health for tracked streams, for example Healthy or Unhealthy
Deletion rate	Percentage of records deleted during imports
Average run time	Average duration of runs, with a note on skipped or unchanged records

See Health review and thresholds for default thresholds and stream overrides.

Charts

Below the KPI cards, the Dashboard shows:

• Imports over time – stacked volume of created, updated, deleted, and failed entities.
• Run outcomes – distribution of succeeded, partial, failed, cancelled, and aborted runs.
• Failures over time – failed records and failed runs across the selected range.
• Streams added / removed – how many streams were added or removed.

Use these charts to correlate spikes in volume with failures or with streams that recently changed.

Health review and thresholds

The stream health widgets show data from recent runs, data freshness, and failure rates. The error table groups the most frequent errors by stage and error code, for example UPSERT or EXTRACT, with codes such as STREAM_ERROR.

Source field health identifies source-data and composite relationship issues:

• Repeated keys – source records that use the same resolved key
• Unresolved parents – composite child records that could not be linked to a parent record

Use the stream's Key resolution and Composite settings to investigate these counts.

Open the related import from Imports when a stream is unhealthy. Use Run history, Imported Data, and the stream configuration to identify the cause. See Imports and Streams.

Health status is based on failure rate, freshness, and whether the last run failed. Select Default thresholds to alter the tenant baseline thresholds. An import or stream can override the thresholds. Leave a field blank to keep the platform default.

Default thresholds for stream health, with the stream health table and error breakdown in the background

Setting	Platform default	Purpose
Amber above failure rate	1%	Warning threshold for failed records
Red above failure rate	10%	Unhealthy threshold for failed records
Amber after no fresh data for	48 h	Warning when a stream has not imported fresh data
Red after no fresh data for	168 h	Unhealthy when a stream has not imported fresh data
A failed last run is red	On	Treat a failed last run as red. Turn this off when occasional whole-run failures are expected and only the rates above decide health.

To override the tenant defaults for one stream, use its Health tab. See Stream health.