Templates & AI Builder

Start from a pre-built template, generate a complete workflow from natural language, or let the AI suggest optimal configurations for each node.

Workflow Templates

Templates are pre-built workflows designed for common automation patterns. Instead of starting from scratch, pick a template that matches your use case and customize it to fit your needs.

Browsing Templates

When you create a new workflow, the template selection modal appears. You can also access it from the editor toolbar at any time.

Category filter — browse by use case: email automation, meeting management, data sync, notifications, and more.
Difficulty level — templates are tagged as Beginner, Intermediate, or Advanced.
Search — type keywords to filter the template list.
Popularity — see which templates are most-used across the platform.

Each template shows a compatibility score based on your existing connections. A template requiring Gmail and Slack shows 100% if you have both connected, 50% with only one, and a prompt to add the missing connection.

Using a Template

Select a template

Click on a template card to preview its structure — the nodes, connections, and a description of what it does.

Click Use Template

The template's nodes and edges are loaded into your editor. The workflow is created in Draft status.

Configure connections

Each action node that requires a connection will need one assigned. Click each node and select the appropriate connection from the dropdown.

Customize parameters

Adjust any parameters — channel names, email filters, time ranges — to match your specific requirements.

Test and activate

Run a test to verify everything works, then set the workflow to Active.

Template Categories

Category	Example Templates
Email Automation	Inbox triage, auto-responder, email-to-Slack forwarder
Meeting Management	Meeting recap generator, auto-schedule follow-ups
Chat & Notifications	Slack alerts for Jira updates, Teams notification pipeline
File & Document	Google Drive backup, document approval workflow
Data & Spreadsheets	Sheet-to-CRM sync, daily report aggregator
Dev & PM	GitHub PR review notifications, Linear issue triage
General Automation	Multi-step approval chain, data enrichment pipeline

AI Workflow Builder

Describe what you want to automate in plain English, and the AI generates a complete workflow — nodes, connections, and initial configurations — ready for you to review and customize.

How It Works

Open the AI Builder

In the template selection modal, switch to the AI Builder tab. Or click the AI icon in the editor toolbar.

Describe your workflow

Type a natural-language description of what you want to automate. Be specific about the trigger, the services involved, and the desired outcome.

Generate

Click Generate. The AI analyzes your description, selects the appropriate nodes and integrations, and assembles a workflow graph.

Review and refine

The generated workflow appears on the canvas. Review the structure, adjust node positions, and fill in any parameters the AI could not determine (like specific channel names or email addresses).

Writing Effective Prompts

Be specific about services — say 'Gmail' not just 'email', say 'Slack #engineering' not just 'send a message.'
Describe the trigger — 'When a new GitHub issue is labeled as bug...' or 'Every morning at 9 AM...'
Specify the outcome — 'post a summary to Slack and update the Jira ticket status to In Progress.'
Mention conditions — 'Only if the email is from @acme.com' or 'If the amount is over $1,000.'
Keep it focused — one workflow per prompt. Complex multi-path automations work better when built in stages.

Iterative refinement

If the first generation isn't quite right, you can use the AI Modify feature to tell the AI what to change: "Add a condition to filter by priority" or "Replace the Slack node with a Teams notification."

Other AI Features

AI Suggest (Node Configuration)

When configuring any node, click the AI Suggest button to let the AI analyze the workflow context and pre-fill configuration values. This is especially useful for complex nodes like AI Prompt (system/user messages) or Transform (field mappings).

AI Modify

After a workflow exists, you can describe changes in natural language. Tell the AI to "add error handling after the API call" or "insert a loop to process each line item." The AI modifies the existing graph rather than regenerating from scratch.

AI Troubleshooting Agent

When an execution fails, the AI Troubleshooting Agent examines the failed node, its input data, connection state, and error message. It produces a diagnosis explaining what went wrong and suggests specific fixes. See the Debugging & Diagnostics guide for details.