Workflows

Buildorado workflows are the core automation engine that powers everything you build on the platform. A workflow is a visual, directed acyclic graph (DAG) of nodes and edges displayed on an infinite canvas, where each node performs a specific task and edges define the order and conditions under which data flows from one step to the next. Whether you are collecting form submissions, routing leads to sales teams, processing payments through Stripe, or generating AI-powered responses, every process you build in Buildorado starts with a workflow.

What Is a Workflow in Buildorado?

Think of a workflow as a flowchart that actually runs. You design it visually on a drag-and-drop canvas, connecting building blocks (nodes) with arrows (edges). When a user submits a form or an API call triggers the workflow, the execution engine processes each node in order, evaluates conditions on each edge, and carries data forward until every path reaches its end.

Unlike simple form builders that only collect data, Buildorado workflows let you act on that data in real time: send personalized emails, charge credit cards, call external APIs, branch based on user answers, loop over datasets, and much more. See the Quick Start guide to create your first workflow in under five minutes.

Workflow Components: Nodes, Edges, and Variables

Every workflow is made up of three core components that work together to define your automation logic.

Nodes: The Building Blocks

Nodes are the individual steps in your workflow. Each node has a single responsibility and produces output data that downstream nodes can consume. Buildorado offers five categories of nodes:

• Form Nodes -- Collect user input with 36 field types including text, email, file upload, signature, payment fields, scheduling, and more. Form nodes are typically the entry point of a workflow.
• Action Nodes -- Execute side effects such as sending emails, posting messages to Slack, creating records in a CRM, triggering webhooks, or collecting payments via Stripe.
• AI Nodes -- Process data using large language models (GPT-4, Claude), generate images, convert speech to text, and perform other AI-powered transformations. Learn more in the AI nodes reference.
• Branch Nodes -- Route execution down different paths based on conditional logic. Support if/else, switch, and multi-output branching with 35+ comparison operators.
• Logic Nodes -- Handle orchestration tasks such as loops, merges, HTTP requests, code execution (JavaScript sandboxed), delays, sub-workflows, data transformations, and error handling. See the full list in the logic nodes reference.

Each node type is covered in detail in the Form Fields, Logic Nodes, and AI Nodes documentation.

Edges: The Connections

Edges are the arrows that connect one node to another. They define both the execution order and, optionally, the conditions under which a path is taken. An edge without conditions is always followed. An edge with conditions is only followed when those conditions evaluate to true based on the data available at runtime.

For example, an edge from a form node to an "Enterprise Sales" action node might carry the condition: budget > 10000 AND company_size > 50. If the submitted data matches, execution continues down that path. If not, the engine evaluates alternative edges. Learn how to configure edge conditions in the Conditional Logic guide.

Variables: The Data Pipeline

Every node produces output variables that downstream nodes can reference. Buildorado uses a double-curly-brace template syntax to inject dynamic values into text fields, email bodies, API payloads, and configurations:

Hello {{contactForm.firstName}}, thanks for your interest in {{contactForm.plan}}!

Each node is assigned a stable variable name when it is created (for example, contactForm, sendEmail_1, httpRequest_2). This name persists even if you rename the node's display label, ensuring that references never break. Variable names follow a pattern of nodeType_sequenceNumber for system and action nodes, while form nodes use the name you assign.

Scoping rules:

• A node can only reference variables from nodes that are upstream in the DAG -- that is, nodes whose output has already been computed by the time the current node executes.
• Form field variables are accessed by their field key: {{formNodeName.fieldKey}}.
• System and action node outputs are accessed by their output schema keys: {{httpRequest_1.statusCode}}, {{agent_1.response}}.
• If a variable name is unambiguous across the entire workflow, you can use the short form without the node prefix.

Workflow Lifecycle: From Draft to Archive

Every workflow moves through a well-defined lifecycle. Understanding each stage helps you organize your work and avoid publishing unfinished automations.

1. Create

Start a new workflow from the dashboard by clicking New Workflow. Choose a blank canvas, start from a template, or describe what you want in plain English and let the AI copilot generate a first draft.

2. Design

Drag nodes onto the canvas, configure their settings, and connect them with edges. Use the left sidebar to browse field types and node categories. Right-click the canvas to access a context menu for adding system nodes, sticky notes, and more.

3. Preview

Click Preview to walk through your workflow exactly as an end user would. The preview mode lets you fill out forms, trigger branches, and verify that actions fire correctly -- all without sending real emails or charging real cards. Preview mode is essential for testing every conditional path before going live.

4. Publish

When you are satisfied with the design, click Publish. The workflow receives a shareable link and can be embedded on your website via an iframe. Published workflows are live and accept real submissions.

5. Monitor

Track submissions, completion rates, and drop-off points from the workflow's analytics dashboard. Each submission generates a detailed execution log showing which nodes fired, what data was passed, and whether any errors occurred.

6. Archive

When a workflow is no longer needed, archive it. Archived workflows stop accepting new submissions but preserve all historical data and execution logs. You can un-archive a workflow at any time to bring it back to a published or draft state.

Workflow States Explained

How the Execution Engine Works

Understanding what happens behind the scenes when a user submits a form helps you design more reliable workflows and debug issues faster.

1. Trigger -- A form submission, API call, webhook, or scheduled event triggers the workflow.
2. DAG Resolution -- The engine computes the execution order by performing a topological sort of the workflow graph. Nodes with no dependencies execute first.
3. Node Execution -- Each node runs in order. Form nodes capture and validate the submission data. Action nodes perform their side effects. AI nodes call external model APIs. System nodes handle orchestration (loops, merges, delays).
4. Edge Evaluation -- After a node completes, the engine evaluates the conditions on every outgoing edge. Only edges whose conditions are satisfied (or edges with no conditions) are followed. This is how branching and routing work.
5. Variable Context -- As each node completes, its output is added to the execution context. Downstream nodes can reference any upstream variable using the {{nodeName.key}} template syntax.
6. Async Processing -- Long-running tasks such as API calls, AI model inference, and file processing are handled asynchronously. The end user is not blocked while these operations complete.
7. Error Handling -- If a node fails, the engine checks for an error handler node attached to it. If one exists, execution is routed to the error handler. If not, the failure is logged and the execution stops on that branch (other branches continue independently).
8. Completion -- The workflow execution is marked as complete when all branches have either finished successfully or been halted by errors.

Version History and Undo/Redo

Buildorado automatically tracks every change you make on the canvas. The version history system provides two levels of protection:

• Undo/Redo -- Use Ctrl+Z / Cmd+Z to undo and Ctrl+Shift+Z / Cmd+Shift+Z to redo any action on the canvas, including adding nodes, deleting edges, moving elements, and changing configurations. The undo stack is preserved for the duration of your editing session.
• Version Snapshots -- Each time you publish a workflow, a snapshot of the current state is saved. You can view previous versions, compare changes, and roll back to any published version if needed.

Collaboration Features

Workflows belong to a workspace, and every member of the workspace can view and edit any workflow within it. Buildorado supports role-based permissions:

• Owner -- Full control, including billing and workspace settings.
• Admin -- Can create, edit, publish, and archive workflows. Can manage integrations and API keys.
• Editor -- Can create and edit workflows but cannot publish or manage integrations.
• Viewer -- Read-only access to workflows and submission data.

Changes made by different team members are tracked in the workflow's activity log, so you always know who modified what and when.

Real-World Example: Lead Qualification Workflow

Here is a step-by-step walkthrough of a lead qualification workflow that collects prospect information, scores the lead, and routes it to the appropriate sales team.

Step 1: Contact Form (Form Node) Create a form with fields for name, email, company name, company size (select: 1-10, 11-50, 51-200, 201-1000, 1000+), annual budget (number), and industry (select).

Step 2: AI Lead Scoring (AI Node) Connect the form to an AI node that uses GPT-4 to analyze the submission and assign a lead score from 1 to 100 based on company size, budget, and industry fit.

Step 3: Branch by Score (Branch Node) Add a branch node with three outgoing edges:

• Score 80 or above: route to Enterprise Sales
• Score 40 or above: route to SMB Sales
• Score below 40: route to Self-Serve Onboarding

Step 4: Enterprise Sales Path (Action Nodes)

• Send a Slack notification to the #enterprise-leads channel with the lead's details.
• Create a deal in your CRM via a webhook.
• Send a personalized email to the prospect: Hi {{contactForm.firstName}}, our enterprise team will reach out within 24 hours.

Step 5: SMB Sales Path (Action Nodes)

• Add the lead to a Google Sheet for the SMB team to review.
• Send an automated email with a link to book a demo via Google Calendar.

Step 6: Self-Serve Path (Action Nodes)

• Send a welcome email with links to documentation and a free trial signup.

This entire workflow runs automatically every time someone submits the contact form. No manual intervention required.

Best Practices for Building Workflows

• Name your nodes descriptively. Instead of "Form 1" and "Action 2", use names like "Contact Form" and "Send Welcome Email". This makes the canvas easier to scan and variable references easier to understand.
• Test every path in preview mode. Before publishing, walk through each branch of your workflow to verify that conditions evaluate correctly and actions fire as expected.
• Use error handler nodes. Attach error handlers to critical action nodes (payment processing, API calls) so that failures are caught and handled gracefully rather than silently swallowed.
• Keep workflows modular. If a workflow grows beyond 15-20 nodes, consider breaking it into sub-workflows. This improves readability and lets you reuse common patterns across multiple workflows.
• Use sticky notes for documentation. Right-click the canvas and add sticky notes to explain complex logic, document business rules, or leave notes for teammates.
• Leverage templates. Browse the template marketplace for pre-built workflows that you can clone and customize rather than starting from scratch.
• Monitor execution logs regularly. Check the analytics dashboard for failed executions, slow nodes, and drop-off points. Address issues before they affect your users.

Frequently Asked Questions

How many nodes can a single workflow contain?

There is no hard limit on the number of nodes per workflow. In practice, workflows with up to 100 nodes perform well. For very large processes, we recommend breaking them into smaller sub-workflows connected via the Sub-Workflow system node for better maintainability and performance.

Can I duplicate or clone a workflow?

Yes. From the dashboard, open the workflow's menu and select Duplicate. This creates an exact copy in draft state, including all nodes, edges, conditions, and configurations. You can also start from a template if you want a pre-built starting point.

What happens to in-progress submissions when I update a published workflow?

When you edit a published workflow, your changes are saved as a new draft. In-progress submissions continue executing against the version that was live when they started. Once you re-publish, only new submissions use the updated version. This ensures that no submission is interrupted mid-execution by a design change.

Can I trigger a workflow without a form submission?

Yes. Workflows can be triggered via incoming webhook, schedule, or from another workflow using a Sub-Workflow node.