Create a Job from a Job Template in the web UI, bind Connectors, run it, and follow the Job Run to completion.

Quickstart: run your first job

Goal

Go from an existing Job Template to a finished Job Run, entirely in the web UI: pick a template, create a Job from it by binding Connectors, run the Job, and watch the run complete.

Prerequisites

A running Envoy instance you can sign in to (First-time setup)
An account with the admin or superadmin role — other roles see the same screens without the run buttons
At least one Job Template in the deployment — customer builds usually ship with templates preloaded
Connectors created for the systems the template needs — see Managing connectors

Steps

1. Open the template

A Job Template is a reusable YAML definition that composes tasks into a DAG and declares which connector types it needs — logic without credentials. See Tasks, jobs, and templates for the model.

Go to Job Templates in the sidebar and select a template from the list. The detail page shows the template's connector slots as badges, its task DAG, and its recent jobs.

2. Create a Job

Select Create Job. The dialog "Configure connectors and arguments to create a runnable job from this template" collects everything the Job needs:

Name (required) — for example production-migration.
Description — optional.
Connector Mapping — one select per template connector slot, filtered to Connectors of the matching type. When exactly one Connector of the right type exists, it is preselected.
Arguments — shown only when the template declares arguments. Required arguments must be filled; optional ones show their default as the placeholder.

Select Create Job in the dialog footer. Envoy snapshots the template, substitutes argument values, stores the connector bindings, and opens the new Job's detail page.

The Job detail page lists the bound Connectors and the template's tasks. To run everything, select Run All in the header. To run a subset, check individual tasks and select Run N Selected; by default upstream depends_on tasks run first, and the Skip dependencies checkbox limits the run to only the checked tasks.

Starting a run navigates you straight to the Job Run page.

4. Follow the Job Run

A Job Run is one execution of the Job's DAG; each task inside it becomes its own task run with its own logs. The Job Run #N page updates live. The header shows the run's status badge, start time, and duration; the table below lists each task in the run with its own status (pending, running, success, failed, cancelled, or skipped).

Select any task row to open that task run's logs and metrics.
Stop (shown while the run is active) cancels the run.
Resume (shown when the run is paused or failed) continues from the incomplete tasks — completed tasks are not repeated.

The Job Run reaches completed when its tasks finish. What happens on a task failure depends on the template's failure policy (pause, abort, or continue) — see Run lifecycle.

Verify it worked

On the Job detail page, the run you started appears under the Job's run history with a completed status badge. Open a task row inside the run and check its logs to confirm records were processed.

Jobs — everything on the Jobs and Job Templates pages, including archiving and export
Job Runs — the run list, statuses, and log viewing in depth
Triggers — run Jobs on a cron schedule instead of by hand
Quickstart: build your first task — the implementer path