Integration Studio and connected applications
Build pages and rules around connected data. Link outside records to the work they affect.
Describe the connected record
An outside data source becomes more useful when its records have a clear meaning in DART. Integration Studio defines the kinds of records supplied by an API. An API is a way for systems to exchange data. A custom record type gives the incoming fields a useful structure.
Typed links connect that data to DART records. For example, a weather alert can link to a duty station. The source record stays available behind the link.
A type gives fields their required values and formats. Objects retain their source and version. A link can relate two records while preserving each record’s identity and protection.
One custom type can combine records from approved APIs. Field ownership remains explicit, so a local action cannot rewrite an API-owned field. Source changes reach those fields through a checked import.
Place data in the operational view
Operators need to see outside information beside the work it affects. A custom page chooses which fields to show and how to sort them. Connected data can appear beside an incident or person. It can also appear on an equipment or station page.
Staff publish an import after checking and applying it. Settings define how old data may be before it is treated as stale. Updates can be paused when needed.
A page must belong to the published connection revision and be enabled for use. Editing settings does not make the new revision live. Readers still need access to each selected object and linked native record.
Configured views can present records as tables, maps, charts, or cards. The chosen fields and geometry must exist in the typed data. A published view does not supply a missing source value.
Rules that create operational work
Some source changes need attention without waiting for someone to notice them on a page. A rule checks a mapped value against a condition. For example, it can test whether a water level is above a set limit.
The result can create a task or inbox notice for named people. A trigger decides whether it runs once or when a value changes. A limit-crossing trigger acts when the reading passes the boundary.
Rules can combine conditions, require a condition to hold, or wait through a cooldown. A held condition needs another committed observation; a timer alone does not create fresh source evidence.
A crossing trigger must be armed again by the reset condition before firing another crossing. Recipients need access to both the connection and the linked work before a notice or task is created.
Controlled custom actions and outbound sends
A connected workflow may need to send a request back to another system. Designers can create custom actions and API send actions. Sends are queued with their payload and result. Staff can inspect the history.
Connections support OAuth sign-in. They can also receive signed webhooks or event streams. The selected provider supplies the required credentials and access rights.
An action must be listed on a published page for the correct object type. The selected object version is checked, and a repeated request returns its saved action result only when the input matches.
Record matches can be linked, rejected, or undone against their current fingerprint. Retrying an outbound delivery requires current settings and a destination marked safe to retry; Failed and Unclear do not automatically mean no outside effect occurred.
A page action can create a manual record without an existing selection. Typed steps can create, update, link, tag, archive, or restore records. A later step failure rolls back the action’s local changes.
Weather and water-level workflows
A worked example helps show how capture, mapping, and rules fit together. The help tasks include a National Weather Service alert connection. Training guides also include a worked NOAA integration. These examples follow the data from capture to use.
A water-level field can be converted from meters to feet. A link places it beside the relevant record. A rule can create a task when the reading crosses a chosen level.
The worked source remains visible behind the mapped value and task. A unit conversion, record link, rule observation, and completed task each leave separate evidence, so the operator can inspect where the result came from.