A workflow is a visual representation of a series of tasks organized in a logical order to achieve a specific goal. Each task is shown as a node, and workflows are designed to automate processes, streamline activities, and respond effectively to different events.
Key characteristics of workflows include:
- Inputs: Datasets utilized in your workflow, datasets inside your project or external APIs.
- Nodes: Individual components representing specific tasks or actions within the workflow.
- Output: The result generated by your workflow (optional). It can be a dataset stored in a project or Workspace datasets.
How to build a Workflow?
Here's a comprehensive guide to building a workflow using Atlas:
1. Create a new workflow
Begin by entering a project and navigate to the Workflow mode in the left side navigation.
Then click on New workflow
You can start from scratch or select a suitable template closest to the idea you are building.
2. Add workflow input
Datasets that you want to use in your workflow.
Drag and drop a input components from the right panel or just drag the dataset from the left layers panel directly.
3. Add logic nodes
Logic nodes perform tasks and actions within your workflow.
Logic nodes let you apply filters, run calculations, transform data, and actions.
Drag them in from the right panel and connect them to your inputs. You can chain multiple nodes to build up your logic.
Think of each node as a step in your process. Want to clean data? Filter it? Calculate something? This is the place.
4. Add workflow output
Optionally, include an output node at the end of your workflow.
You can return the output data in a dataset that is stored in the project or in Workspace datasets.
5. Run workflow
Then click on the Run button to run the workflow.
Congratulations! You've successfully created a dynamic workflow tailored your needs.
The visual canvas editor
The workflow editor is a drag-and-drop canvas. The right panel lists every available node type grouped by category (data, filters, geometry, raster, AI tools, connectors). Drag nodes onto the canvas and connect them by dragging from a node's output handle to another node's input. Connections enforce type compatibility — you can't connect a raster output into a vector input.
Click any node to select it; its settings appear in a side panel. Selected nodes can be deleted, duplicated, or repositioned. The canvas supports zoom, pan, and a minimap for orienting in larger workflows.
You can run the entire workflow with the Run button, or right-click an individual node to run only up to that point — useful when iterating on a single step.
Workflow templates
Workflow templates let you start from a curated configuration instead of a blank canvas. When you create a new workflow, the Templates modal shows a gallery of common patterns (e.g., "Buffer + dissolve + count", "Travel time isochrones", "Spatial join with attribute filter"). Pick a template to drop a fully wired-up workflow onto the canvas, then customise it for your data.
Templates are read-only originals — using one creates a copy, so the source template is never mutated. You can also save your own workflow as a template for the workspace by selecting Save as template from the workflow's options menu.
Scheduling and triggering
Workflows can run on demand, on a schedule, or in response to an event. See Scheduling, Webhook trigger, and POST request for the trigger mechanisms.