Dashboard

App page

The Dashboard is the workspace home page. Use it to resume work, see what is running, and open common actions without browsing the list pages first.

Where to find it

Select Dashboard in the main menu. It is also the page you land on after signing in.

On a narrow window or a zoomed page, the main menu, Jobs and Runtime Executions included, sits behind the Menu button at the left of the top bar.

The main menu marks the page you are on. Each item is a link, so it opens in a new tab from the browser's link menu.

What you see

Layout

The page reads top to bottom: header, an optional setup banner, the KPI ribbon, then a two-column body. The left column holds Activity, Environments and Projects. The right column holds Scheduled, AI credits and Data moved.

• Finish setting up your workspace lists the remaining setup tasks with a Continue setup action. It disappears once every required step is complete. Apply Permission is listed as Optional and does not count toward the total. Nine steps are required.
• The header is titled Dashboard, with a greeting by first name under it.
• The KPI ribbon shows Executions today, Success rate, Active jobs, Connection health and Total projects. The Connection health icon is green when every connection passed its last test, amber while any is untested, and red when any failed. A tile with nothing to measure yet reads None.
• Activity shows the execution trend chart with the latest runs underneath, in one card.
• Scheduled lists what runs next.
• Projects lists your recent, pinned and all projects behind three tabs.
• AI credits shows this cycle's AI usage against the monthly budget.
• Data moved shows records read from sources and written to targets over the last 7 days.
• Environments shows one tile per environment.

The environment chips in the header filter the schedule to one environment. Refresh reloads the dashboard data.

Each card loads on its own. When you return to the dashboard, cards show their last figures at once and read Updating while newer figures load.

Activity

• Switch between Total, Success and Failed.
• Switch the chart style: Line, Area or Spline.
• Toggle Labels to show or hide the count at each data point.

Under the chart, Latest runs has one row per recently active project, showing that project's most recent run and its status. View all opens the full execution history.

Scheduled

A live countdown to the next run, then up to five jobs due in the next 24 hours, each with its time, project, environment and cadence.

When nothing is due in the next 24 hours, the card shows the next few runs beyond that window, with the day against each one. The header says which of the two lists you are looking at.

Projects

Three tabs over the same rows. Every row shows the project name, whether it is deployed or still a draft, when it was last updated, the environment its jobs run in, and a pin control.

Tab	Shows
Recent	Up to five, yours first. Projects you opened recently lead, and the account's most recently updated fill any remaining rows.
Pinned	Only pinned projects.
All	Every project, newest first, capped at five rows. The number on the tab is the account total.

A project with no jobs shows no environment. Projects are not assigned to an environment themselves; the environment is read from their jobs.

AI credits

Everyone on the account sees the usage bar and the per-surface breakdown. The dollar figures are shown to Admins and Owners only.

The card shows tokens used instead of a percentage in three cases, and says which one applies:

• No AI usage yet this cycle.
• No monthly budget is set.
• Cost is unavailable for some calls. One such call withholds the account's whole dollar figure.

The breakdown is a share of tokens, not of cost, so it still reports in that last case.

The graph underneath shows tokens per day over the last 30 days. The figures above it cover the current billing month, so the two do not add up.

The card needs the Owner or Admin role.

Data moved

Records read from sources and written to targets over the last 7 days, shown as two figures with a daily trend and the change against the previous 7 days.

• Pulled counts the records your source nodes read; Written counts the records your target nodes wrote. They are separate figures: a filter reads more than it writes, and adding the two would count a plain copy twice.
• Each figure reports its own change against the previous 7 days. No change is shown when the previous period had no runs.
• Moved by project ranks the projects that moved the most records in the window.
• An account with no recorded volume sees "No records moved yet".
• If some runs did not report their row counts, the total is marked as a lower bound.

Size appears under each figure, one for what was pulled and one for what was written.

• It is an estimate of the size of those rows.
• It is not bytes sent over the network. A compressed response or a bulk upload moves fewer bytes than this figure.
• Written can exceed pulled for the same rows, because a target's rows carry the result columns the write adds.
• A run that did not report a size shows none, rather than 0 B. When only some runs in the window reported one, the figure covers those runs only.

View runs opens the execution history.

Environments

One tile per environment, with its health, its scheduled-run counts for the last 24 hours, and the project behind any problem. Manage opens the Environments page.

The health pill covers the runtime, not connections. It reads the status and heartbeat of the runtime linked to the environment:

Pill	Meaning
Healthy	The runtime is running and its heartbeat is current.
Heartbeat slow	Running, last heartbeat over 5 minutes ago.
Heartbeat stale	Running, last heartbeat over 30 minutes ago.
Degraded	The runtime is stopped, failed or unreachable.
Unknown	No runtime is linked to this environment.

For connection health, open the Environments page.

These are the same words the Environments page uses. A failing data pipeline does not change them: run health and environment health are separate, and an environment whose runtime is fine reads Healthy even while a job in it fails.

The counts are scheduled runs. A run you start by hand from the canvas is not included.

• The second figure is failures. With no failures it shows stuck runs instead, and with neither it shows zero failed.
• A run still going 2 hours after it started is reported as stuck. A job that normally runs longer than that appears here too, with its project named.
• The line underneath names the project behind the failure or the stuck run. With neither, it shows the count of stopped and skipped runs as warnings, or what the health pill means for this environment.
• An environment with no scheduled runs in the window shows zero.

Common tasks

Run or pause the next scheduled job

Both controls on the Scheduled card act on the next job, and both need the same job schedule permission the Jobs page uses.

• Run now queues the job to start immediately, the same way the scheduler does, so the run appears in the runs list and the Environments counts. The card says Run queued: the run has been accepted, not finished.
• Pause turns the schedule off. A paused job has no next run, so it leaves this card, and a Resume button stays in its place. Resuming can be refused on plans that do not allow the schedule, and the card says so.

Pin a project

Select the pin control on a project row. Pinning holds up to eight projects. Pinning the same one again unpins it. The Pinned tab lists them.

Hide or restore the setup banner

The close button on Finish setting up your workspace hides the banner for you alone, not for your colleagues. It asks for a second click. The checklist stays available under Manage Account > Team Access > Workspace setup, which is also where you turn the banner back on.

Related

• Projects, Jobs and Environments.
• Runtime Executions: the full run history.